[{"data":1,"prerenderedAt":1416},["ShallowReactive",2],{"helpPage-advice":3,"projects-advice":70,"blogs-advice":323,"testimonial-advice":1254,"servicePage-index-page":1269},{"id":4,"title":5,"blockDescription":6,"blockIcon":7,"blocks":8,"body":21,"description":40,"extension":44,"h1":45,"h2":46,"insights":47,"insightsTitle":48,"largeCoverImage":49,"meta":50,"navigation":53,"path":54,"projects":55,"seo":60,"stem":67,"testimonial":68,"__hash__":69},"help\u002Fhelp\u002Fadvice.md","Technology Consultancy","Offering strategic advice, technology consultancy and support on end-to-end digital transformation.","\u002Fimg\u002Fmisc\u002Fservice-groups\u002Ftechnology-path-white.svg",[9,12,15,18],{"title":10,"description":11},"Technology","We'll work together to analyse the current state of processes across your business, consulting around what value driven solutions are a good fit for now, and the future, based on your short and long term goals.",{"title":13,"description":14},"People","We'll dive into team setups, including who is responsible for delivering each digital transformation initiative, as well as providing consultancy and support around team development, engagement and retention.",{"title":16,"description":17},"Strategy","We'll collaborate to define your business vision, including long term and short term objectives, helping us to deliver an appropriate technology roadmap that identifies new opportunities for growth and innovation, with solutions that are robust, secure and scalable.",{"title":19,"description":20},"Delivery","We'll help you to implement the correct methodologies and processes throughout your business, such as agile as an iterative approach to transformation projects, with the appropriate levels of stakeholder engagement to ensure success in delivery.",{"type":22,"value":23,"toc":39},"minimark",[24,29,33,36],[25,26,28],"h2",{"id":27},"empowering-businesses-to-create-the-foundations-for-growth-and-gain-a-competitive-edge","Empowering businesses to create the foundations for growth and gain a competitive edge",[30,31,32],"p",{},"Digital transformations are in full effect. Companies globally are collaborating with partners to improve performance, find new revenue streams, and provide compelling customer experiences that support their goals.",[30,34,35],{},"We're a leading team of business analysts and technology consultants with extensive experience in strategy and delivery expertise across industries, providing vendor agnostic technology advice to find value driven solutions that remove complexities and scale.",[30,37,38],{},"We work with you to understand what you are trying to achieve in the short and long term, aligning a robust technology strategy that sets out a roadmap to your target operating model, ensuring that solutions are secure and scalable, delivering value to your customers and achieving long term success.",{"title":40,"searchDepth":41,"depth":41,"links":42},"",2,[43],{"id":27,"depth":41,"text":28},"md","Offering strategic consultancy and support on end-to-end digital transformation","Providing technology consultancy, insight and support to uncover new opportunities for growth and innovation, ensuring you have the correct processes, projects and people in place to deliver digital success.",null,"Insights","\u002Fimg\u002Fconsulting.jpg",{"bgColour":51,"projectsTitle":52},"#0c457d","Projects",true,"\u002Fhelp\u002Fadvice",[56,57,58,59],"specialist-marine-consultants-smc","adm-agriculture","o2","aesseal",{"title":61,"description":62,"keywords":63,"jsonSchema":64},"Digital Transformation Consulting | Audacia","Digital transformation consultancy, Audacia, offers strategic advice and support on end-to-end digital transformation for leading brands.","Digital transformation consulting, digital transformation consultancy, technology consulting, technology consultancy, software consulting, software consultancy, software development consulting, software development consultancy",[65],{"json":66},"{\n  \"@context\": \"https:\u002F\u002Fschema.org\u002F\", \n  \"@type\": \"BreadcrumbList\", \n  \"itemListElement\": [{\n    \"@type\": \"ListItem\", \n    \"position\": 1, \n    \"name\": \"Home\",\n    \"item\": \"https:\u002F\u002Faudacia.co.uk\u002F\"  \n  },{\n    \"@type\": \"ListItem\", \n    \"position\": 2, \n    \"name\": \"Technology Consultancy\",\n    \"item\": \"https:\u002F\u002Faudacia.co.uk\u002Fhelp\u002Fadvice\"  \n  }]\n}","help\u002Fadvice","tom-broadbent-aesseal-plc","F-bRsqQEWWZn8u5GVjdn1cpks0-8Rq3w7-XinxaC_tI",[71,185,258],{"id":72,"title":73,"about":74,"bgColor":47,"body":75,"clientLogo":160,"coverImageColour":161,"description":40,"extension":44,"h1":47,"h2":47,"image":47,"insightsTitle":47,"keyTech":162,"largeCoverImage":163,"meta":164,"navigation":53,"pageSections":47,"path":168,"projectName":169,"projectState":170,"projectStatement":47,"projectsTitle":47,"results":171,"seo":178,"smallCoverImage":47,"smallTitle":47,"statBlocks":47,"stem":183,"__hash__":184},"projects\u002Fprojects\u002Fadm-agriculture.md","ADM Agriculture","ADM Agriculture is a UK subsidiary of ADM, one of the world’s largest agricultural processors and food ingredient providers, with more than 31,000 employees, serving customers in 170+ countries. \n",{"type":22,"value":76,"toc":154},[77,81,84,87,91,94,97,100,104,107,110,113,116,119,122,125,132,136,139,142,145,148,151],[25,78,80],{"id":79},"client","Client",[30,82,83],{},"ADM Agriculture is a UK subsidiary of ADM, one of the world’s largest agricultural processors and food ingredient providers, with more than 31,000 employees, serving customers in 170+ countries.",[30,85,86],{},"ADM Agriculture operates from offices throughout England and offers a people-based, quality service to farmers and consumers. As well as providing an integrated supply chain to ADMs UK assets in milling and oil seed crushing, ADM Agriculture supplies a full range of non-grain feed ingredients to the feed and fuel markets, alongside a comprehensive range of seed and fertilisers.",[25,88,90],{"id":89},"background","Background",[30,92,93],{},"ADM Agriculture’s legacy trading platform was becoming a barrier to the business’ growth, with many employees’ jobs starting to involve implementing workarounds to the system to carry out their roles, costing the business time and resources.",[30,95,96],{},"ADM sought to find a software development company that could develop a modern commodities trading platform as part of a complex legacy upgrade. It was important that the platform could scale in line with future business growth plans and make time savings for staff.",[30,98,99],{},"The company manages three core product areas: fertiliser, seed and grain. Each has 100's of different products and pricing rules, so ADM Agriculture was looking for a partner who could translate complex business logic into an intuitive and secure system. Building a fully bespoke system would give the company the flexibility and speed required to manage the complex workflows that they were dealing with on a daily basis. ",[25,101,103],{"id":102},"solution","Solution",[30,105,106],{},"Audacia and ADM Agriculture worked together through analysis sessions to understand granular levels of detail of business processes, define scenarios and develop and test prototypes to create a fully future-proof system. This initial analysis discovered how employees were carrying out their work and honed in on what the business wanted to achieve. ",[30,108,109],{},"Collaborating with stakeholders and end-users throughout the development process, Audacia used a range of technologies, including SQL Server, Azure DevOps, .NET Core, Typescript, Angular and Cypress, to build a modern commodities trading platform. ",[30,111,112],{},"Integrated with ADM Agriculture’s two other core systems, NetSuite and Red Tractor Assurance, the platform streamlines and automates key areas of the business including purchase and sales contracts for products, pricing structures on contracts (added as fixed process or variable price dependent on market value at time of execution), and integrated price alerts to notify traders on price fluctuations.",[30,114,115],{},"The system contains vital functionality to support the business-critical invoicing module, streamlining the process of generating over 107,000 invoices a year. Covering deliveries between farms and consumers, including invoices for the farmer, haulier and consumer, with digitised footprints for claims made on delivery if the product does not match expected quality.",[30,117,118],{},"Providing synchronised and accurate information, there is consistent data transfer across the platform, updating factors including invoice generation and supplier validity in real-time, minimising the costs and obstructions to the commodity trading process and improving financial return.",[30,120,121],{},"Audacia worked closely with ADM Agriculture stakeholders to develop internal training programmes on the system and continue to support ADM as they work to further optimise business functions through integrated applications.  ",[30,123,124],{},"Audacia helped ADM transition to an agile project management methodology and, working in collaboration with the company’s Head of IT, helped implement these processes across the organisation. Through these changes, ADM was able to maintain ongoing scalability and flexibility, supporting the project's evolving requirements and ensuring KPIs were met.",[30,126,127],{},[128,129],"img",{"alt":130,"src":131,"title":130},"ADM software development project screens","\u002Fimg\u002Fscreens\u002Fadm-screens.png",[25,133,135],{"id":134},"results","Results ",[137,138],"project-results",{":results":134},[30,140,141],{},"The solution meets the immediate needs of ADM today and has the capability to scale in line with future business plans. The project met all KPIs, budget and timescales and it is completely scalable, which allows for future changes and growth of the company.  The platform itself is user-focused, eliminating workarounds to the previous system as well as providing significant time and cost savings across the organisation.",[30,143,144],{},"The ADM trading platform is currently responsible for over £3.5 billion in trading activity with a global rollout currently underway. In addition, the new system has eliminated workarounds, as well as leading to a 97% reduction in processing time for cancelling truck movements and 12x faster invoicing speeds.",[30,146,147],{},"As a result of the platform's flexibility and scalability, the platform has been successfully rolled out across several of ADM Agriculture’s subsidiaries.",[30,149,150],{},"With the successful launch of this core trading platform, ADM has continued to invest in innovative technology solutions that can support their future growth plans. Examples of recent innovations include mobile applications for harvest sampling and a web-based portal for all customer-facing work. This additional portal allows customers to review sales contracts, delivery orders, balances over time and store collections, while farmers can review purchase contracts (grain), sales contracts (seed\u002Ffertiliser) and collections.",[30,152,153],{},"ADM Agriculture and Audacia have established a long-term partnership based on the quality of results secured and the strength of the client relationship developed through the project.",{"title":40,"searchDepth":41,"depth":41,"links":155},[156,157,158,159],{"id":79,"depth":41,"text":80},{"id":89,"depth":41,"text":90},{"id":102,"depth":41,"text":103},{"id":134,"depth":41,"text":135},"\u002Fimg\u002Fclient-logos\u002Fadm.png","#f6e2c2","ASP.NET Core, Angular, Octopus Release Management, Azure DevOps, Microsoft SQL Server, C#\n","https:\u002F\u002Fimages.unsplash.com\u002Fphoto-1565647952915-9644fcd446a4?ixlib=rb-1.2.1&ixid=MnwxMjA3fDB8MHxwaG90by1wYWdlfHx8fGVufDB8fHx8&auto=format&fit=crop&w=2070&q=80",{"statBlock":165},{"title":166,"text":167},"£3.5 billion","Commodity contracts and services supported for one of the world's largest agricultural organisations","\u002Fprojects\u002Fadm-agriculture","A commodities trading platform to support £3.5bn contracts and services each year","In Development",[172,174,176],{"text":173},"10.5 million MT of commodity traded per year",{"text":175},"£3.5 billion contracts and services supported annually",{"text":177},"International rollout across 16 countries",{"title":179,"keywords":180,"description":181,"image":182},"Commodities Trading Platform - ADM Agriculture","Commodities trading platform, Trading software, Bespoke software, Bespoke software development, Software development company, Agriculture software development","Software development company, Audacia, worked with global agricultural leader, ADM, to develop a bespoke commodities trading platform.","\u002Fimg\u002Fscreens\u002Faudaciaadmsoftwaredevelopment.jpg","projects\u002Fadm-agriculture","6WMwxdpn38vRbW_-aRYBD02TZjMyXjtZ0XptQQ6dqpM",{"id":186,"title":187,"about":188,"bgColor":47,"body":189,"clientLogo":233,"coverImageColour":234,"description":40,"extension":44,"h1":47,"h2":47,"image":47,"insightsTitle":47,"keyTech":235,"largeCoverImage":236,"meta":237,"navigation":53,"pageSections":47,"path":241,"projectName":242,"projectState":243,"projectStatement":47,"projectsTitle":47,"results":244,"seo":251,"smallCoverImage":47,"smallTitle":47,"statBlocks":47,"stem":256,"__hash__":257},"projects\u002Fprojects\u002Fo2.md","O2","O2 is the commercial brand of Telefónica UK Limited and is a leading digital communications company with the highest customer satisfaction for any mobile provider. With over 25 million customers, O2 runs 2G, 3G and 4G networks across the UK, as well as operating O2 Wifi and owning half of Tesco Mobile.\n",{"type":22,"value":190,"toc":227},[191,193,203,205,208,210,213,219,222,224],[25,192,80],{"id":79},[30,194,195,196,202],{},"O2 is the commercial brand of Telefónica UK Limited and is a leading digital communications company with the highest customer satisfaction for any mobile provider. With over 25 million customers, O2 runs 2G, 3G and 4G networks across the UK, as well as operating O2 Wifi and owning half of Tesco Mobile. ",[197,198,187],"a",{"href":199,"rel":200},"https:\u002F\u002Fwww.o2.co.uk\u002Fandroid-tablets",[201],"nofollow"," has over 450 retail stores and sponsors The O2, O2 Academy venues and England Rugby.",[25,204,90],{"id":89},[30,206,207],{},"Previously, Telefonica were using excel spreadsheets to hold their data and had a manual based process, consisting of emailing ticket availability to recipients. The process had no auditing control, which presented a risk factor of human error. In order to create a more seamless and robust process and improve efficiency, O2 looked for a software development company that was experienced in delivering reliable and cost effective bespoke software solutions.",[25,209,103],{"id":102},[30,211,212],{},"Working in partnership with Wrights, a UK based marketing agency, Audacia developed a web based platform to be used by account managers, partner managers, team managers and event administrators at O2. The bespoke software application was developed to manage all aspects of the business process; from budgeting tickets for events, through to issuing tickets to customers.",[30,214,215],{},[128,216],{"alt":217,"src":218,"title":217},"O2 software development project screens","\u002Fimg\u002Fscreens\u002Fo2-screens.png",[25,220,221],{"id":134},"Results",[137,223],{":results":134},[30,225,226],{},"The intuitive platform was very well received, significantly improving their business processes. The use of the new software application resulted in an increase in efficiency, a greater robust process and made the whole ticketing process much simpler for O2.",{"title":40,"searchDepth":41,"depth":41,"links":228},[229,230,231,232],{"id":79,"depth":41,"text":80},{"id":89,"depth":41,"text":90},{"id":102,"depth":41,"text":103},{"id":134,"depth":41,"text":221},"\u002Fimg\u002Fclient-logos\u002Fo2.png","#254252","ASP.NET, SQL Server, Email Marketing, Excel Integration\n","\u002Fimg\u002Fo2-cover-image.jpg",{"statBlock":238},{"title":239,"text":240},"Simplifying ","Ticketing processes.","\u002Fprojects\u002Fo2","Managing end-to-end ticketing processes","Released",[245,247,249],{"text":246},"Simplified ticketing process",{"text":248},"Intuitive, user friendly design",{"text":250},"Improved efficiency, reduced errors",{"title":252,"keywords":253,"description":254,"image":255},"Bespoke Web-Based Ticketing Software - O2","Bespoke ticketing software, Ticketing platform, Bespoke ticketing platform, Event ticketing system, Software development company, Bespoke web application, Bespoke web application development, Integrated web application","Software development company Audacia partnered with O2 to develop a bespoke web based application to manage all aspects of their ticketing & administration process.","\u002Fimg\u002Fscreens\u002Faudaciao2softwaredevelopment.jpg","projects\u002Fo2","3Kl7eDkMI-Gk7I-GYcHeisc8vwb7lkjM2IFiDs42Ias",{"id":259,"title":260,"about":261,"bgColor":47,"body":262,"clientLogo":298,"coverImageColour":299,"description":40,"extension":44,"h1":47,"h2":47,"image":47,"insightsTitle":47,"keyTech":300,"largeCoverImage":301,"meta":302,"navigation":53,"pageSections":47,"path":306,"projectName":307,"projectState":47,"projectStatement":47,"projectsTitle":47,"results":308,"seo":315,"smallCoverImage":47,"smallTitle":320,"statBlocks":47,"stem":321,"__hash__":322},"projects\u002Fprojects\u002Fspecialist-marine-consultants-smc.md","Specialist Marine Consultants (SMC)","Specialist Marine Consultants (SMC) are a global leader in offshore solutions. SMC work with the biggest global companies involved in oil and gas exploration and offshore renewable energy, having deployed their skilled HSEQ Advisors on international marine projects across every continent and ocean.\n",{"type":22,"value":263,"toc":292},[264,266,269,271,274,276,279,285,287,289],[25,265,80],{"id":79},[30,267,268],{},"Specialist Marine Consultants (SMC) are a global leader in offshore solutions - providing HSEQ and marine consultancy, marine coordination, offshore project management, client representation, vessel inspections and bespoke training. SMC work with the biggest global companies involved in oil and gas exploration and offshore renewable energy, having deployed their skilled HSEQ Advisors on international marine projects across every continent and ocean.",[25,270,90],{"id":89},[30,272,273],{},"As SMC continued to expand, they began to look for a critical technology partner to redevelop their legacy systems and introduce new functionality. This redevelopment would ensure their further growth and further improve the business' efficiency by allowing intelligent management of offshore and onshore assets, vessels and personnel transfers. After researching UK-based software development companies, SMC decided to engage with Audacia due to our proven track record in business-critical development projects.",[25,275,103],{"id":102},[30,277,278],{},"Audacia’s development team worked in partnership with SMC to develop a bespoke online portal that allows marine co-ordinators to manage assets at sea, such as turbines, substations and vessels, including the construction of turbines, with real-time visibility of the status of each assets; for example, current problems, exclusion zones and build status, in one centralised platform. Alongside assets, the platform also enables SMC to effectively execute the daily planning of personnel between locations for different job tasks, such as building, maintenance or review. In managing personnel transfers, the platform also contains reporting functions for factors such as distance travelled, fuel consumption, passenger transfer and number of 'push on's'.",[30,280,281],{},[128,282],{"alt":283,"src":284,"title":283},"SMC software development project screens","\u002Fimg\u002Fscreens\u002Fsmc-screens.png",[25,286,221],{"id":134},[137,288],{":results":134},[30,290,291],{},"The new bespoke software platform enables SMC to facilitate the optimal transfer of personnel with the visibility needed to effectively optimise planning (for example, the size of vessels sent out to specific locations). The integrated software solution also allows for greater collaboration between planning staff and onsite technicians to monitor factors such as safety information, including permits to work, certificate validations and RFIs, significantly improving operational efficiencies.",{"title":40,"searchDepth":41,"depth":41,"links":293},[294,295,296,297],{"id":79,"depth":41,"text":80},{"id":89,"depth":41,"text":90},{"id":102,"depth":41,"text":103},{"id":134,"depth":41,"text":221},"\u002Fimg\u002Fclient-logos\u002Fsmc.png","#3c9ad3","ASP.NET, Angular, Azure DevOps, Entity Framework, Typescript\n","https:\u002F\u002Fimages.unsplash.com\u002Fphoto-1548337138-e87d889cc369?ixlib=rb-1.2.1&ixid=MnwxMjA3fDB8MHxwaG90by1wYWdlfHx8fGVufDB8fHx8&auto=format&fit=crop&w=1196&q=80",{"statBlock":303},{"title":304,"text":305},"212,937","Transfers managed across wind farms for an award winning offshore solutions provider","\u002Fprojects\u002Fspecialist-marine-consultants-smc","A platform to manage people across the biggest offshore windfarms in the world",[309,311,313],{"text":310},"Real-time visibility of onshore & offshore assets",{"text":312},"Facilitating the optimal transfer of personnel",{"text":314},"Greater collaboration between departments",{"description":316,"title":317,"keywords":318,"image":319},"Software development company Audacia delivers a platform for marine co-ordinators to manage assets at sea, such as turbines, substations and ships, with real-time updates of all assets.","Legacy System Redevelopment - Specialist Marine Consultants","Legacy system, Legacy system redevelopment, Legacy system modernisation, Software development company, Bespoke software development, Offshore management software, Asset management software, Vessel management software","\u002Fimg\u002Fscreens\u002Faudaciasmcsoftwaredevelopment.jpg","SMC","projects\u002Fspecialist-marine-consultants-smc","f0jqqKMUypoyoQZeuokfWCfCHFucE9Tm-fDSvEp1u9I",[324,622,944,1206],{"id":325,"title":326,"author":327,"blogTags":328,"body":330,"customExcerpt":609,"date":610,"description":334,"excerpt":47,"extension":44,"h1":47,"image":611,"meta":612,"navigation":53,"path":613,"readingStats":614,"seo":619,"stem":620,"__hash__":621},"blog\u002Fblog\u002Fuser-driven-development-in-digital-product-delivery.md","Building What Matters – Reducing the 80% Feature Waste in Product Delivery ","Richard Brown",[329],"Engineering",{"type":22,"value":331,"toc":600},[332,335,338,347,356,359,362,365,368,372,375,378,381,389,392,395,399,407,441,444,452,461,464,467,471,474,477,480,491,494,497,501,504,507,510,513,517,520,528,536,540,543,546,560,563,566,569,576,580,583,589,597],[30,333,334],{},"Half of everything software teams build is wasted.  ",[30,336,337],{},"This is a consistent finding across multiple studies and decades of data, and it should be the starting point for any conversation about digital product delivery. ",[30,339,340,341,346],{},"The Standish Group's foundational CHAOS ",[197,342,345],{"href":343,"rel":344},"https:\u002F\u002Fwww.standishgroup.com\u002Fproducts\u002Fcopy-of-chaos-report-beyond-infinity-digital-version",[201],"research",", tracking feature usage across mission-critical applications, found that 50% of features are used almost never, 30% are used infrequently, and only 20% are used often. ",[30,348,349,350,355],{},"Whilst Pendo's 2024 ",[197,351,354],{"href":352,"rel":353},"https:\u002F\u002Fwww.pendo.io\u002Fresources\u002Fthe-hidden-cost-of-bad-software\u002F",[201],"product benchmarks",", drawing on aggregated usage data from hundreds of software products, states that approximately 80% of features built never achieve meaningful adoption - only 12% of features generate 80% of average daily usage volume.  ",[30,357,358],{},"These numbers mean that the majority of engineering effort on most digital products is spent building things that users do not value enough to use. ",[30,360,361],{},"The financial implications are significant. Across an organisation with multiple product teams, the cumulative waste can run into millions. This cost is also compounded, with unused features adding maintenance burden, increasing testing complexity, creating additional attack surface for security vulnerabilities, and making the product harder to learn and navigate for the users who remain. ",[30,363,364],{},"The question of what to build is therefore the highest-leverage decision in any digital product programme. If the team builds the right things, most of the other decisions become easier. If the team builds the wrong things, it can become increasingly difficult for any amount of technical excellence to recover the investment. ",[30,366,367],{},"This article looks at why technology projects consistently lose focus on user outcomes, what the evidence says about the impact of user-centred delivery, and how engineering leaders can build the discipline of tying every feature to genuine user value. ",[25,369,371],{"id":370},"the-drift-from-user-outcomes","The Drift from User Outcomes ",[30,373,374],{},"Technology projects rarely set out to ignore users. The initial brief is always framed around user needs: we are building this system to help these people do these things more effectively. Early conversations focus on user journeys, pain points, and desired outcomes, with the first sprints typically delivering visible, user-facing functionality. ",[30,376,377],{},"Drift can also happen driven by technical. As the project progresses, the backlog accumulates technical work, such as infrastructure setup, security hardening, database optimisation, API refactoring, dependency upgrades, and performance tuning, with each task being individually justified and often genuinely necessary. The problem is that these tasks are framed in technical language, disconnected from any user-facing outcome, and prioritised alongside (or above) user stories without a shared framework for comparison. ",[30,379,380],{},"This can lead to sprint reviews starting to demonstrate infrastructure work to stakeholders who have no way to assess whether the project is on track, because nothing they can see or experience has changed. The feedback loop that keeps the product aligned with user needs gradually weakens and can eventually break. The team is still delivering, stories are completed, velocity is maintained, deployment pipelines are green, but what is being delivered has drifted from the purpose the project was commissioned to serve. ",[30,382,383,384,388],{},"PMI's 2025 Pulse of the Profession ",[197,385,345],{"href":386,"rel":387},"https:\u002F\u002Fwww.pmi.org\u002Flearning\u002Fthought-leadership\u002Fboosting-business-acumen",[201],", surveying nearly 3,000 project professionals, found that only 14% of employees feel aligned with the goals of their organisation. When this misalignment exists at an organisational level, it compounds within product delivery teams. Engineers build what the backlog tells them to build. Product owners prioritise what stakeholders request. Stakeholders request what they believe is important, often informed by the loudest customer complaint, the most recent executive conversation, or a competitor feature they noticed, rather than by systematic analysis of user behaviour. Without a disciplined connection to actual user outcomes, each layer of this chain can introduce drift. ",[30,390,391],{},"The stakeholder request pattern is a particularly common source of wasted effort. Stakeholders often translate user problems into solution specifications: \"we need a PDF export button\" rather than \"users need to share report data with colleagues who do not have system access.\" The first framing prescribes a specific solution, whereas the second framing describes a user need that could be addressed in multiple ways, some of which may be simpler, faster and more effective than the solution initially assumed. When the team builds the specified solution without interrogating the underlying need, they risk building something that technically works but does not solve the problem the user actually has. ",[30,393,394],{},"Structural factors can also compound this drift. In many organisations, contracts define success by scope delivered rather than outcomes achieved, creating an incentive to build everything specified regardless of whether it is needed. Project sponsors request broad functionality to avoid perceived gaps, whilst legacy systems set a precedent for feature coverage, with modernisation teams cautious about removing functionality even where usage is low or unclear, and roadmaps are built to demonstrate activity rather than reflect evidence of use. Each of these drivers shapes delivery priorities away from user focus, and in environments without consistent user validation, completed features accumulate even when they serve no meaningful purpose. ",[25,396,398],{"id":397},"the-evidence-for-user-centred-delivery","The Evidence for User-Centred Delivery ",[30,400,401,402,406],{},"Forrester's widely cited ",[197,403,345],{"href":404,"rel":405},"https:\u002F\u002Fwww.forrester.com\u002Freport\u002Fthe-six-steps-for-justifying-better-ux\u002FRES117708",[201]," suggests that every £1 invested in user experience design yields a return of £100 – a 9,900% ROI. Other studies suggest similar: ",[408,409,410,414,423,432],"ul",{},[411,412,413],"li",{},"The Interaction Design Foundation cites a similar rule of thumb: every £1 invested in UX saves £10 in development and £100 in post-release maintenance.  ",[411,415,416,417,422],{},"The Baymard Institute ",[197,418,421],{"href":419,"rel":420},"https:\u002F\u002Fbaymard.com\u002Flearn\u002Fux-statistics",[201],"attributes"," high rates of drop-off and conversion loss in transactional systems to usability failures, and consistent user input has been shown to improve decision-making and delivery outcomes across multiple studies. ",[411,424,425,426,431],{},"McKinsey's Business Value of Design ",[197,427,430],{"href":428,"rel":429},"https:\u002F\u002Fwww.mckinsey.com\u002Fcapabilities\u002Fmckinsey-design\u002Four-insights\u002Fthe-business-value-of-design",[201],"study",", tracking 300 publicly listed companies over five years, found that organisations in the top quartile for design maturity achieved 32% higher revenue growth and 56% higher total shareholder returns.  ",[411,433,434,435,440],{},"Whilst The Design Management Institute’s Design Value Index ",[197,436,439],{"href":437,"rel":438},"https:\u002F\u002Fwww.dmi.org\u002Fpage\u002FDesignValue",[201],"found"," that design-led companies outperformed the S&P 500 by 219% over a ten-year period.  ",[30,442,443],{},"These studies demonstrate that companies which consistently pay attention to how users interact with their systems tend to perform better overall. ",[30,445,446,447,451],{},"PMI's 2025 ",[197,448,345],{"href":449,"rel":450},"https:\u002F\u002Fwww.pmi.org\u002Flearning\u002Fthought-leadership\u002Fproject-success",[201]," adds a more recent data point. The report found that high-performing project professionals track an average of 9.1 success factors per project, compared to 6.3 for others. Critically, the additional factors these professionals track include customer satisfaction, strategic alignment, and stakeholder trust – measures  that go well beyond the traditional iron triangle of scope, schedule, and budget. The PMI data found that organisations prioritising interpersonal capabilities, including collaborative leadership, communication and empathy, achieve 72% business goal success rates, compared to 65% for those that do not. Empathy, in a project context, means understanding what users actually need and building accordingly. ",[30,453,454,455,460],{},"Gartner ",[197,456,459],{"href":457,"rel":458},"https:\u002F\u002Fwww.gartner.com\u002Fen\u002Fnewsroom\u002Fpress-releases\u002F2019-02-19-gartner-survey-finds-85-percent-of-organizations-favor-a-product",[201],"reports"," that approximately 85% of organisations now favour a product-centric application delivery model, reflecting a broad industry shift toward continuous, outcome-oriented development and away from project-based delivery with fixed scope and predetermined features. ",[30,462,463],{},"This raises an obvious question: if the vast majority of organisations now favour product-centric delivery, why does feature waste remain so high? The answer is that favouring a model and operating one are very different things. Many organisations have adopted the language and structures of product thinking, such as product owners, product teams, and outcome-based roadmaps, while the underlying behaviours remain project-oriented. ",[30,465,466],{},"The evidence base is consistent in that user-centred delivery produces better outcomes by every meaningful measure: adoption, revenue, retention, satisfaction and return on investment. ",[25,468,470],{"id":469},"product-thinking-in-practice","Product Thinking in Practice ",[30,472,473],{},"The discipline of keeping features focused on users is fundamentally about product thinking. This is an orientation where features are treated as experiments with hypotheses: “We believe that building this capability will deliver this outcome for this user group”. Each sprint delivers something a user can see, or respond to, which creates the feedback loop that keeps the product aligned with actual needs. ",[30,475,476],{},"This stands in contrast to project thinking, where features are treated as deliverables on a timeline: “we committed to delivering this module by this date”. Project thinking measures progress by output (stories completed, features shipped). Product thinking measures progress by outcome (user adoption, task completion rates, satisfaction scores, time saved). The distinction shapes everything from backlog prioritisation to how success is measured. ",[30,478,479],{},"The principle is straightforward: every piece of work should link to a deliverable feature or component that serves a user need. There will always be lower-level technical requirements, such as infrastructure, security, and performance optimisation, and these are essential – the discipline is in how they are framed. For example: ",[408,481,482,485,488],{},[411,483,484],{},"The authentication refactor exists so that users can log in reliably with single sign-on ",[411,486,487],{},"The caching layer exists so that the dashboard loads in under two second ",[411,489,490],{},"The API redesign exists so that the mobile app can display real-time data ",[30,492,493],{},"This framing keeps the team oriented toward outcomes and gives stakeholders meaningful progress signals. ",[30,495,496],{},"When technical work cannot be connected to a specific user-facing capability, it is worth questioning whether it needs to happen now, or at all. Some technical work is genuinely foundational and has no immediate user-visible output. Infrastructure, security hardening, compliance requirements, and resilience improvements all fall into this category – they are essential to the product's viability even though users may never interact with them directly. This is acceptable, provided the team can articulate which user-facing capability it enables and when that capability will be demonstrable. A backlog where infrastructure work is connected to upcoming user stories maintains alignment. A backlog where infrastructure work accumulates without a clear path to user value is a question point. ",[25,498,500],{"id":499},"the-sprint-demo-as-diagnostic","The Sprint Demo as Diagnostic ",[30,502,503],{},"There is a simple process for whether a project has maintained its focus on user outcomes. At the end of every sprint, can the team demonstrate something to a user (or a convincing proxy for a user) and get meaningful feedback? ",[30,505,506],{},"If the answer is consistently yes, the project has a functioning feedback loop where stakeholders can see progress in terms they understand, and users can respond to real functionality, providing the kind of concrete feedback that written requirements can never capture. This means that the team can course-correct early, adjusting priorities based on what they learn from demonstrating working software. ",[30,508,509],{},"If the answer is consistently no, and sprints are consumed by technical work that is invisible to the people the product is being built for, then the project has lost its connection to user outcomes. This does not necessarily mean the work is wasted, but it does mean that the primary mechanism for validating whether the team is building the right thing has been suspended. Every sprint without user feedback can become a sprint where assumptions go untested and drift goes undetected. ",[30,511,512],{},"However, this is not to say that every sprint is required to deliver a polished, user-facing feature. Some sprints will focus on enabling infrastructure, and that is expected. The discipline is in maintaining the cadence of user-visible progress. If three consecutive sprints pass without anything demonstrable to a user, the team should discuss whether this is justified and plan when the next user-visible delivery will occur. ",[25,514,516],{"id":515},"the-hidden-costs-technical-debt-and-rework","The Hidden Costs: Technical Debt and Rework ",[30,518,519],{},"The cost of building the wrong features extends well beyond the initial wasted engineering effort. Two downstream consequences, technical debt and rework, compound the damage in ways that are often invisible until they become severe. ",[30,521,522,523,527],{},"Technical debt is typically attributed to shortcuts in architecture or code quality, but research shows it is also shaped by delivery choices, particularly around unnecessary features. When teams implement functionality that adds limited user value, they increase the system's size, surface area and maintenance cost. A ",[197,524,430],{"href":525,"rel":526},"https:\u002F\u002Fwww.sciencedirect.com\u002Fscience\u002Farticle\u002Fpii\u002FS0167642318301035",[201]," published in ScienceDirect found that 25% of engineering effort goes toward managing technical debt – a figure that grows as unused features accumulate in the codebase. Every feature that remains in the system, whether used or unused, must be maintained, tested, secured and accounted for in future architectural decisions. The system becomes harder to understand, change and extend, even when the code quality of each individual feature is high. ",[30,529,530,531,535],{},"Rework is an equally significant cost. Further ",[197,532,345],{"href":533,"rel":534},"https:\u002F\u002Fwww.sciencedirect.com\u002Fscience\u002Farticle\u002Fabs\u002Fpii\u002FS0950584913002188",[201]," published in ScienceDirect found that 40-50% of development effort is spent on rework, often because features did not align with user needs or because requirements shifted after implementation began. Rework becomes less expensive when it happens early – teams that gather and act on user input during discovery and prototyping reduce the likelihood of revising features after release. This is a further argument for the continuous user validation that sprint demos and embedded product ownership provide: the cost of learning that a feature misses the mark in a sprint review is a fraction of the cost of discovering the same misalignment after deployment. ",[25,537,539],{"id":538},"the-prioritisation-discipline","The Prioritisation Discipline ",[30,541,542],{},"The feature usage data provides the strongest argument for rigorous prioritisation. If 80% of features go unused, or 50% are used almost never, then building every feature that stakeholders request, with equal weight and priority, virtually guarantees that the majority of engineering effort will be wasted. ",[30,544,545],{},"The remedy for this is to tie every feature to a user outcome with a testable hypothesis. Before a feature enters the backlog, the team should be able to articulate:  ",[408,547,548,551,554,557],{},[411,549,550],{},"which users will benefit,  ",[411,552,553],{},"what they will be able to do that they cannot do today,  ",[411,555,556],{},"how the team will know if the feature is successful (through measurable adoption, task completion or satisfaction), and  ",[411,558,559],{},"what the cost of building and maintaining this feature will be relative to its expected value. ",[30,561,562],{},"This discipline can be uncomfortable. It requires potentially saying no to stakeholders who want features built, accepting that some ideas which seem compelling in a workshop will not survive contact with evidence, product owners to defend prioritisation decisions with data, and engineering leaders to support them in doing so. It also requires the organisational maturity to treat features as investments that must justify their cost, and to retire features that fail to deliver value, rather than allowing them to accumulate indefinitely. ",[30,564,565],{},"The accumulation problem can be seen in most digital products only ever adding features. Features that were built, launched, found little adoption and quietly stopped being useful remain in the product, consuming maintenance effort, increasing test suite size, adding complexity to the codebase and confusing users who encounter them. The discipline of feature retirement – actively removing functionality that is not delivering value – is one of the most neglected practices in product development, and one of the most valuable. Every feature removed is maintenance effort recovered, test complexity reduced and user experience simplified. ",[30,567,568],{},"This issue becomes especially visible during modernisation efforts. Teams are understandably cautious about removing legacy functionality, even where usage is low or unclear. Carrying features forward feels safer than evaluating their current value. This can be aided by bringing evidence to decision making. Factors such as usage analytics, service logs and user feedback can identify which legacy behaviours remain genuinely important. Where data is not available, teams can flag features for conditional review during prototyping, introduce them only when a user need is confirmed and capture workarounds or support cases as signals of feature relevance. This approach supports leaner, more maintainable systems that evolve in response to current needs rather than historical defaults. ",[30,570,571,572,575],{},"Pendo's ",[197,573,345],{"href":352,"rel":574},[201]," found that product adoption and usage now rank as the most important metrics of product success among product leaders – a marked change from prior years, when product leaders were more likely to measure success by features shipped. The industry is moving, slowly, from measuring output to measuring outcomes. Organisations that make this shift at the project level and tie every feature to a user outcome and measuring whether it delivered, should result in wasting far less of their engineering investment. ",[25,577,579],{"id":578},"keeping-the-user-in-the-room","Keeping the User in the Room ",[30,581,582],{},"The most effective mechanism for maintaining user focus is to keep the user (or a credible representation of the user) in the room throughout the project. This means embedded product ownership from someone who deeply understands user needs and has the authority to make prioritisation decisions, regular user research as a continuous practice throughout delivery, analytics on existing features informing what to build next, and using real usage data to validate assumptions and identify where users are struggling. ",[30,584,446,585,588],{},[197,586,345],{"href":449,"rel":587},[201]," found that projects with a clear vision of success, defined in terms of value delivered, score +41 on PMI's Net Project Success Score, compared to -18 for those without one. A clear vision describes the value the product will deliver to the people who use it, and it provides the reference point against which every feature decision can be evaluated. When the vision is clear, prioritisation becomes easier because every feature can be assessed against a shared understanding of what the product is trying to achieve and for whom. ",[30,590,591,592,596],{},"The technology industry has spent decades improving how software is built – better languages, frameworks, infrastructure, testing, and deployment pipelines. These improvements are genuine and valuable. However, the persistent gap in project success rates suggests that the industry has invested heavily in building things right while underinvesting in building the right things. With ",[197,593,595],{"href":449,"rel":594},[201],"50%"," of projects globally now classified as successful (a figure that has barely improved despite enormous advances in delivery capability, the remaining gap is overwhelmingly about what teams choose to build, not how they build it. ",[30,598,599],{},"User-centred feature delivery addresses this gap directly, and the evidence that it works is among the most robust in the entire field of software engineering. The organisations that close this gap will be those that bring the same rigour to deciding what to build that they already apply to how they build it – treating features as investments to be validated, measured and, when necessary, retired.",{"title":40,"searchDepth":41,"depth":41,"links":601},[602,603,604,605,606,607,608],{"id":370,"depth":41,"text":371},{"id":397,"depth":41,"text":398},{"id":469,"depth":41,"text":470},{"id":499,"depth":41,"text":500},{"id":515,"depth":41,"text":516},{"id":538,"depth":41,"text":539},{"id":578,"depth":41,"text":579},"Half of everything software teams build is wasted. This is a consistent finding across multiple studies and decades of data, and it should be the starting point for any conversation about digital product delivery. This article looks at why technology projects consistently lose focus on user outcomes, what the evidence says about the impact of user-centred delivery, and how engineering leaders can build the discipline of tying every feature to genuine user value. ","2026-04-29T13:54:52.396Z","\u002Fimg\u002Fblog\u002Fuser-driven-development.png",{},"\u002Fblog\u002Fuser-driven-development-in-digital-product-delivery",{"text":615,"minutes":616,"time":617,"words":618},"14 min read",13.69,821400,2738,{"title":326,"description":334},"blog\u002Fuser-driven-development-in-digital-product-delivery","7eLff1s8F_IJ9Rj63mPTisiTzusI2Z8T3VqRHLEBeMw",{"id":623,"title":624,"author":625,"blogTags":626,"body":629,"customExcerpt":931,"date":932,"description":633,"excerpt":47,"extension":44,"h1":47,"image":933,"meta":934,"navigation":53,"path":935,"readingStats":936,"seo":941,"stem":942,"__hash__":943},"blog\u002Fblog\u002Fdata-foundations-for-scalable-ai.md","Enterprise Data Foundations: The Determinant of AI at Scale ","Ryan Crompton",[627,628],"AI","Data",{"type":22,"value":630,"toc":906},[631,634,637,640,667,670,674,677,680,689,693,696,701,704,718,721,725,728,732,735,739,742,746,749,752,756,759,762,765,768,771,780,788,792,795,798,802,805,809,812,816,819,822,826,829,832,835,838,849,852,856,865,869,872,876,879,883,886,890,893,897,900,903],[30,632,633],{},"Every enterprise AI conversation eventually becomes a conversation about data. The question is whether it happens early enough. ",[30,635,636],{},"The AI initiative begins with an exciting use case, a promising model and a confident timeline. Weeks or months later, the project stalls – not because the model doesn't work, but because the data it depends on is incomplete, inconsistent, inaccessible or untrustworthy. The team spends months on bespoke data engineering, the timeline slips, the business case erodes and the pilot joins the growing catalogue of promising experiments that never reached production.  ",[30,638,639],{},"This is the single most common structural failure in enterprise AI: ",[408,641,642,650,658],{},[411,643,644,649],{},[197,645,648],{"href":646,"rel":647},"https:\u002F\u002Fwww.gartner.com\u002Fen\u002Fnewsroom\u002Fpress-releases\u002F2025-02-26-lack-of-ai-ready-data-puts-ai-projects-at-risk",[201],"Gartner"," predicts that through 2026, organisations will abandon 60% of AI projects unsupported by AI-ready data.  ",[411,651,652,653,657],{},"A ",[197,654,656],{"href":646,"rel":655},[201],"survey"," of 248 data management leaders found that 63% of organisations either do not have, or are unsure if they have, the right data management practices for AI.  ",[411,659,660,661,666],{},"Cisco's 2024 AI ",[197,662,665],{"href":663,"rel":664},"https:\u002F\u002Fnewsroom.cisco.com\u002Fc\u002Fr\u002Fnewsroom\u002Fen\u002Fus\u002Fa\u002Fy2024\u002Fm11\u002Fcisco-2024-ai-readiness-index-urgency-rises-readiness-falls.html",[201],"Readiness Index",", assessing nearly 8,000 organisations globally, found that 80% report inconsistencies or shortcomings in data pre-processing and cleaning for AI projects. ",[30,668,669],{},"The message is that data readiness is the determinant that separates organisations that scale AI from those who stay in permanent experimentation. ",[25,671,673],{"id":672},"why-traditional-data-management-falls-short","Why Traditional Data Management Falls Short ",[30,675,676],{},"Most enterprise data estates were not built for AI. They were built for reporting and transactions – structured databases, ETL pipelines feeding data warehouses, quality rules focused on completeness and consistency for dashboards and regulatory returns. These systems serve their original purpose well, but they are insufficient for the demands AI places on data. ",[30,678,679],{},"Traditional data management optimises for structured data, batch processing and known query patterns. AI-ready data management must additionally handle unstructured data at scale (documents, images, audio, free text), support real-time or near-real-time data integration, enable feature engineering and storage for machine learning pipelines, provide robust data lineage and provenance tracking, accommodate the iterative, experimental nature of model development and balance accessibility with governance – making data discoverable and usable without compromising security or compliance. ",[30,681,682,683,688],{},"A lack of necessary data to train effective AI models was ",[197,684,687],{"href":685,"rel":686},"https:\u002F\u002Fwww.rand.org\u002Fpubs\u002Fresearch_reports\u002FRRA2680-1.html",[201],"identified"," as the second most common root cause of project failure, based on interviews with 65 experienced data scientists and engineers. As reported, executives often believe they have great data because they receive weekly sales reports, without realising that data serving one purpose may be wholly inadequate for another. ",[25,690,692],{"id":691},"the-five-dimensions-of-ai-data-readiness","The Five Dimensions of AI Data Readiness ",[30,694,695],{},"Data readiness for AI is a capability that can be assessed across several dimensions: ",[697,698,700],"h3",{"id":699},"quality","Quality  ",[30,702,703],{},"Quality goes well beyond traditional measures such as accuracy and completeness, for AI, quality also encompasses:  ",[408,705,706,709,712,715],{},[411,707,708],{},"representativeness (does the data reflect the real-world conditions the model will encounter in production, including edge cases and demographic diversity?);  ",[411,710,711],{},"timeliness (is data current enough for the use case – critical for real-time applications, less so for historical analysis?); ",[411,713,714],{},"consistency (are the same concepts measured the same way across systems?); and  ",[411,716,717],{},"label quality (for supervised learning, are the labels accurate, consistent and unbiased?).  ",[30,719,720],{},"Poor data quality can reduce model accuracy but can also introduce systematic biases that are difficult to detect and costly to correct. ",[697,722,724],{"id":723},"accessibility","Accessibility  ",[30,726,727],{},"Accessibility determines whether data can be discovered, accessed and used by AI teams without weeks of negotiation with data owners, manual extraction processes or informal workarounds. In many organisations, the data exists but is locked in silos, subject to unclear ownership, accessible only through legacy systems with limited APIs, or governed by policies written before AI was a consideration. Organisations need to build collaborative, cross-domain strategies for data access as they move from AI pilots to operational AI. ",[697,729,731],{"id":730},"integration","Integration  ",[30,733,734],{},"Integration addresses the ability to combine data from multiple sources reliably and repeatably. Most valuable AI use cases require joining data across systems, such as customer data with transactional records, operational data with external signals, or structured data with unstructured content. Each integration point introduces complexity, such as schema mismatches, temporal misalignment, and identity resolution challenges. Without automated, repeatable integration pipelines, AI projects can become a manual data wrangling exercise. ",[697,736,738],{"id":737},"governance","Governance  ",[30,740,741],{},"Governance ensures that data usage for AI complies with regulatory requirements, organisational policies and ethical standards. This includes data privacy and consent management, data lineage (tracing how data moves through the organisation and into models), access controls and the ability to audit how data was used in training and inference. As the regulatory landscape evolves – the EU AI Act now applying obligations for general-purpose AI models since August 2025 – governance is becoming no longer optional for any AI initiative. ",[697,743,745],{"id":744},"architecture","Architecture  ",[30,747,748],{},"Architecture determines whether the data infrastructure can support AI workloads at scale. This encompasses compute and storage capacity for model training and inference, real-time data streaming capabilities, feature stores that allow engineered features to be shared across models and teams and the overall design of the data platform – centralised, federated or hybrid. ",[30,750,751],{},"An assessment across these five dimensions will reveal where the gaps are, which can often be bigger than initially assumed.  ",[25,753,755],{"id":754},"the-hidden-tax-why-data-debt-compounds","The Hidden Tax: Why Data Debt Compounds ",[30,757,758],{},"The cost of inadequate data foundations can ultimately result in every AI project becoming a bespoke data engineering exercise, and the cost of that approach compounds over time. ",[30,760,761],{},"A typical pattern can be:  ",[30,763,764],{},"An AI team is tasked with building a customer churn prediction model. They spend three months sourcing, cleaning, integrating and preparing the data – work that is specific to this use case, this data combination and this team's particular workarounds for the organisation's data quality issues. The model performs well and the pilot is deemed a success. ",[30,766,767],{},"Six months later, a different team is tasked with building a customer propensity model. They need much of the same underlying data – customer demographics, transaction history, engagement metrics – but the previous team's data preparation work is undocumented, unreproducible or inaccessible. So, they start again from scratch. Another three months of data engineering, with another set of bespoke pipelines and another set of quality workarounds. ",[30,769,770],{},"Each project bears the full cost of data preparation, with none of that investment reusable for the next initiative. Over time, the organisation accumulates not a data platform but a tangle of disconnected pipelines, each serving a single use case, each maintained (or not) by a different team. The marginal cost of the next AI initiative never decreases, there is no compounding benefit and the data estate becomes progressively harder to govern, audit and secure. ",[30,772,773,774,779],{},"This is often ",[197,775,778],{"href":776,"rel":777},"https:\u002F\u002Fwww.mckinsey.com\u002Fcapabilities\u002Fmckinsey-digital\u002Four-insights\u002Fcharting-a-path-to-the-data-and-ai-driven-enterprise-of-2030",[201],"referred to"," as the “pilot purgatory\" trap: teams across the enterprise launch proof-of-concept models that have no chance of scaling because they were built on one-off data foundations that cannot support production deployment. The enthusiasm around generative AI has intensified this pattern, as stakeholders invest in use cases that each require building entire data architectures before value can be realised. ",[30,781,782,783,787],{},"BCG's ",[197,784,345],{"href":785,"rel":786},"https:\u002F\u002Fwww.bcg.com\u002Fpublications\u002F2024\u002Fwheres-value-in-ai",[201]," displays the alternative, with AI leaders allocating roughly 20% of their resources to technology and data foundations as an ongoing strategic investment, as opposed to a one-off investment. This investment in shared infrastructure is what allows them to pursue fewer, more focused AI initiatives while achieving more than twice the ROI of their peers. The data platform is the compounding asset and without it, every initiative starts from zero. ",[25,789,791],{"id":790},"architecture-patterns-that-support-scalable-ai","Architecture Patterns That Support Scalable AI ",[30,793,794],{},"The choice of data architecture determines how quickly new AI use cases can move from idea to production, how much each initiative costs and whether the organisation can govern and scale its AI capabilities across the enterprise. ",[30,796,797],{},"Several architectural patterns have emerged as particularly relevant for AI-ready data environments, each with distinct strengths. ",[697,799,801],{"id":800},"datalakehouse","Data lakehouse  ",[30,803,804],{},"Data lakehouse architectures combine the flexibility of data lakes (handling structured, semi-structured and unstructured data) with the governance and performance features of data warehouses. They support both traditional analytics and machine learning workloads on a single platform, reducing the duplication and complexity that comes from maintaining separate systems for different use cases. For organisations seeking a middle ground, lakehouses offer scalability, SQL compatibility and increasingly native ML features such as feature stores and vector search. ",[697,806,808],{"id":807},"data-fabric","Data fabric  ",[30,810,811],{},"Data fabric architectures take a metadata-driven approach, creating an intelligent integration layer across disparate data sources without requiring data to be physically centralised. The fabric uses metadata – information about the data itself – to automate data discovery, governance, quality monitoring and access across cloud, on-premise and hybrid environments. This approach is particularly valuable in regulated sectors where data cannot easily be moved or consolidated, and where governance and lineage are non-negotiable. Gartner has positioned data fabric as a foundational architecture for AI-ready enterprises. ",[697,813,815],{"id":814},"data-mesh","Data mesh  ",[30,817,818],{},"Data mesh represents an organisational shift as much as a technical shift. Rather than centralising data management, data mesh distributes ownership to domain teams (marketing owns marketing data, finance owns financial data), while a central team provides shared infrastructure, governance standards and self-service tooling. Each domain treats its data as a product, with clear ownership, quality standards and documentation. The approach can work well for large, complex organisations with decentralised structures, provided they have the governance maturity to maintain enterprise-wide standards across federated teams. ",[30,820,821],{},"In practice, most organisations will adopt hybrid approaches, with three broad archetypes identified – centralised, decentralised and hybrid – with the choice being driven by business objectives and consumption needs, not by technology preference alone. The critical principle is that the architecture should make data discoverable, accessible, trustworthy and reusable across AI initiatives, not just optimised for a single use case. ",[25,823,825],{"id":824},"a-path-forward","A Path Forward ",[30,827,828],{},"The most common objection to investing in data foundations is that it sounds like a multi-year, enterprise-wide programme that must be completed before any AI value can be delivered. However, the organisations that succeed do not necessarily choose between \"fix all the data first\" and \"ignore data and build pilots.\", they instead look to a third path: building data foundations iteratively, aligned to the highest-priority AI use cases, with each investment creating reusable assets that accelerate the next initiative. ",[30,830,831],{},"The approach has several elements: ",[30,833,834],{},"Start with the use case portfolio rather than the data estate ",[30,836,837],{},"Rather than attempting to catalogue and clean all enterprise data, identify the two or three highest-priority AI initiatives, using outcome-first prioritisation, and assess the specific data requirements for each.  ",[408,839,840,843,846],{},[411,841,842],{},"What data sources are needed?  ",[411,844,845],{},"What quality standards must be met?  ",[411,847,848],{},"What integration is required?  ",[30,850,851],{},"This focuses the data investment on what matters most, not on abstract completeness. ",[697,853,855],{"id":854},"conductadatareadiness-assessment","Conduct a data readiness assessment ",[30,857,858,859,864],{},"For each priority use case, evaluate data across the five dimensions: quality, accessibility, integration, governance and architecture. This assessment should involve both data teams and business stakeholders – the latter often have critical context about data quality issues that technical assessments miss. ",[197,860,863],{"href":861,"rel":862},"https:\u002F\u002Fwww.dataversity.net\u002Farticles\u002Fdata-management-trends\u002F",[201],"61%"," of organisations list data quality as a top challenge, but only 11% have high metadata management maturity, showing that the gap between recognising the problem and acting on it remains the central challenge. ",[697,866,868],{"id":867},"build-shared-reusable-data-products","Build shared, reusable data products ",[30,870,871],{},"Where multiple AI use cases require the same underlying data (they often do), invest in building that data as a shared product: well-documented, quality-assured, governed and accessible through standard interfaces. A unified customer data product, for example, can serve churn prediction, propensity modelling, personalisation and lifetime value analysis – rather than each team building their own version from raw source data. This is the investment that can create compounding returns. ",[697,873,875],{"id":874},"embed-governance-from-the-start","Embed governance from the start ",[30,877,878],{},"Data governance should be designed in from day one, including data lineage, access controls, quality monitoring, consent management and audit capabilities. Organisations moving from AI pilots to operational AI need collaborative, cross-domain governance strategies that span data management, model management and risk management. ",[697,880,882],{"id":881},"invest-in-data-engineering-capacity","Invest in data engineering capacity ",[30,884,885],{},"Data engineering is the discipline that transforms raw data into AI-ready assets, and it is chronically under-resourced relative to data science in most organisations. Without sufficient data engineering capacity, data scientists spend the majority of their time on data preparation rather than model development – a well-documented and persistent pattern. ",[697,887,889],{"id":888},"iterate-and-expand","Iterate and expand ",[30,891,892],{},"The data platform is an evolving capability. Each AI initiative should leave the data estate in better shape than it found it, including new pipelines documented, quality improvements captured, and governance controls extended. Over time, this iterative approach builds a comprehensive, AI-ready data foundation without requiring a “big-bang” transformation programme. ",[25,894,896],{"id":895},"conclusion","Conclusion ",[30,898,899],{},"Data foundations are the foundations of the AI-enabled enterprise – invisible when working, catastrophic when failing. ",[30,901,902],{},"The organisations that get the data part right will build a compounding capability in a data estate that becomes more valuable, more accessible and more AI-ready with every initiative. Resulting in each new use case being faster and cheaper to deliver than the last, each shared data product serving multiple applications, and each governance improvement reducing risk across the portfolio. ",[30,904,905],{},"Investing in data readiness should be framed as the accelerator that makes AI possible at scale, as opposed to the prerequisite that delays it.",{"title":40,"searchDepth":41,"depth":41,"links":907},[908,909,917,918,923,930],{"id":672,"depth":41,"text":673},{"id":691,"depth":41,"text":692,"children":910},[911,913,914,915,916],{"id":699,"depth":912,"text":700},3,{"id":723,"depth":912,"text":724},{"id":730,"depth":912,"text":731},{"id":737,"depth":912,"text":738},{"id":744,"depth":912,"text":745},{"id":754,"depth":41,"text":755},{"id":790,"depth":41,"text":791,"children":919},[920,921,922],{"id":800,"depth":912,"text":801},{"id":807,"depth":912,"text":808},{"id":814,"depth":912,"text":815},{"id":824,"depth":41,"text":825,"children":924},[925,926,927,928,929],{"id":854,"depth":912,"text":855},{"id":867,"depth":912,"text":868},{"id":874,"depth":912,"text":875},{"id":881,"depth":912,"text":882},{"id":888,"depth":912,"text":889},{"id":895,"depth":41,"text":896},"Most enterprise AI projects stall not because the model fails, but because the data underneath is incomplete or inaccessible. This article sets out the five dimensions of AI data readiness, examines how data debt compounds across initiatives, and explores the architectural patterns that allow organisations to scale AI.","2026-04-29T13:44:21.144Z","\u002Fimg\u002Fblog\u002Fdata-foundations-for-ai.png",{},"\u002Fblog\u002Fdata-foundations-for-scalable-ai",{"text":937,"minutes":938,"time":939,"words":940},"10 min read",9.56,573600,1912,{"title":624,"description":633},"blog\u002Fdata-foundations-for-scalable-ai","TdcwulQyDUNQgSG-QRWY77cm0cKDgWIODqymXeYeZrM",{"id":945,"title":946,"author":47,"blogTags":947,"body":950,"customExcerpt":1193,"date":1194,"description":954,"excerpt":47,"extension":44,"h1":47,"image":1195,"meta":1196,"navigation":53,"path":1197,"readingStats":1198,"seo":1203,"stem":1204,"__hash__":1205},"blog\u002Fblog\u002Fwhat-ai-assisted-engineering-means-for-software-testing.md","What AI-Assisted Engineering Means for Software Testing ",[627,948,949],"Testing","Software Testing ",{"type":22,"value":951,"toc":1178},[952,955,958,961,964,968,971,980,983,986,1000,1003,1012,1015,1024,1028,1031,1039,1047,1050,1053,1056,1059,1062,1066,1069,1072,1080,1083,1086,1090,1093,1097,1100,1104,1107,1111,1114,1118,1121,1125,1128,1132,1135,1138,1141,1144,1147,1151,1154,1157,1165,1168,1172,1175],[30,953,954],{},"In February 2025, Andrej Karpathy coined the term \"vibe coding\" to describe a development practice where you \"fully give in to the vibes, embrace exponentials, and forget that the code even exists.\" The developer describes intent in natural language; the AI generates code; the developer runs it, provides feedback, and iterates, often without reading or understanding the code being produced. Collins English Dictionary named it Word of the Year for 2025 and by the end of that year, it had moved from a whimsical observation about weekend projects to a defining challenge for engineering leadership. ",[30,956,957],{},"This practice has evolved beyond hobby projects and hackathons. Microsoft's CEO has disclosed that up to 30% of the company's code is now AI-generated, with Google reporting similar figures, Y Combinator reported that 25% of startups in its Winter 2025 batch had codebases that were 95% AI-generated, and over 90% of developers now report using AI coding tools in their workflows.  ",[30,959,960],{},"The spectrum ranges from responsible AI-assisted development, where AI augments an engineer who reviews, tests and takes ownership of all generated code, to pure “vibe coding”, where code is accepted uncritically in pursuit of speed.  ",[30,962,963],{},"The question for engineering leaders is not whether their teams are using AI to write code (they almost certainly are), but whether the testing, review and quality practices surrounding that code have evolved to match the pace and nature of AI-assisted development.  ",[25,965,967],{"id":966},"the-quality-evidence","The Quality Evidence ",[30,969,970],{},"The data on AI-generated code quality has matured rapidly during 2025, moving from anecdotal concerns to large-scale empirical research.  ",[30,972,973,974,979],{},"CodeRabbit's State of AI vs Human Code Generation ",[197,975,978],{"href":976,"rel":977},"https:\u002F\u002Fwww.coderabbit.ai\u002Fblog\u002Fstate-of-ai-vs-human-code-generation-report",[201],"report",", published in December 2025, analysed 470 real-world open-source GitHub pull requests – 320 AI-co-authored and 150 human-only – using a structured issue taxonomy: ",[30,981,982],{},"AI-generated pull requests contained 1.7 times more issues overall, averaging 10.83 findings per PR compared with 6.45 for human-authored code ",[30,984,985],{},"AI PRs contained 1.4 times more critical issues and 1.7 times more major issues. ",[408,987,988,991,994,997],{},[411,989,990],{},"Logic and correctness errors 1.75 times more frequently; ",[411,992,993],{},"Code quality and maintainability issues 1.64 times more frequently; ",[411,995,996],{},"Security findings 1.57 times more frequently; and ",[411,998,999],{},"Performance problems 1.42 times more frequently. ",[30,1001,1002],{},"At the 90th percentile, AI pull requests reached 26 issues per change – more than double the human baseline. ",[30,1004,1005,1006,1011],{},"The security dimension is particularly stark. Veracode's 2025 GenAI Code Security ",[197,1007,1010],{"href":1008,"rel":1009},"https:\u002F\u002Fwww.veracode.com\u002Fresources\u002Fanalyst-reports\u002F2025-genai-code-security-report\u002F",[201],"Report"," analysed 80 coding tasks across more than 100 large language models and found that AI-generated code introduced security vulnerabilities in 45% of cases. These were not minor issues and included OWASP Top 10 vulnerabilities. Findings include: Java was the riskiest language with a 72% security failure rate, while Python, C#, and JavaScript logged failure rates between 38% and 45%. Defences against cross-site scripting failed in 86% of relevant samples. ",[30,1013,1014],{},"Perhaps the most concerning finding from Veracode's research is that this is not a problem that improves with better models. Security performance remained flat regardless of model size, training sophistication, or release date. Newer, larger models produce more syntactically correct code, but not more secure code. The models learn from vast public code repositories that contain both secure and insecure patterns, and they reproduce insecure approaches with the same confidence as secure ones. This means that organisations waiting for the next model release to solve their AI code quality problem may be waiting indefinitely. ",[30,1016,1017,1018,1023],{},"A December 2025 ",[197,1019,1022],{"href":1020,"rel":1021},"https:\u002F\u002Fwww.csoonline.com\u002Farticle\u002F4116923\u002Foutput-from-vibe-coding-tools-prone-to-critical-security-flaws-study-finds.html",[201],"assessment"," reinforced these findings through testing. It compared five leading AI coding tools – Claude Code, OpenAI Codex, Cursor, Replit, and Devin – by building the same three test applications with each. The result: 69 vulnerabilities across 15 applications, including several rated critical. The tools performed well at avoiding generic, well-known vulnerability patterns like SQL injection, but failed consistently on context-dependent business logic flaws – the kind that require understanding how a workflow should operate, which AI agents currently lack. ",[25,1025,1027],{"id":1026},"the-technical-debt-accelerator","The Technical Debt Accelerator ",[30,1029,1030],{},"Security vulnerabilities are the most immediately dangerous consequence of unchecked AI-generated code, but they are not the only one. The evidence on structural code quality tells a parallel story of rapidly accumulating technical debt. ",[30,1032,1033,1034,1038],{},"GitClear's AI Copilot Code Quality ",[197,1035,345],{"href":1036,"rel":1037},"https:\u002F\u002Fwww.gitclear.com\u002Fai_assistant_code_quality_2025_research",[201],", analysing 211 million changed lines of code from 2020 to 2024 across repositories owned by Google, Microsoft, Meta, and enterprise organisations, found that AI coding assistants are fundamentally changing the composition of code being written, and not for the better. They found:  ",[408,1040,1041,1044],{},[411,1042,1043],{},"Code duplication exploded - blocks with five or more duplicated lines increased eightfold during 2024.  ",[411,1045,1046],{},"Refactoring collapsed - the proportion of changed lines associated with refactoring fell from 25% in 2021 to less than 10% in 2024, a decline of nearly 40%.  ",[30,1048,1049],{},"For the first time in the history of their dataset, copy-pasted lines exceeded moved lines, meaning developers were duplicating code more than they were consolidating it into reusable modules. ",[30,1051,1052],{},"With regards to code churn, the proportion of new code revised within two weeks of its initial commit, also rose significantly, from 3.1% in 2020 to 5.7% in 2024. This indicates that AI-generated code is being corrected or reworked at higher rates, suggesting that initial output quality is lower even when the code appears to function correctly on first execution. ",[30,1054,1055],{},"The mechanism behind these trends is that AI coding assistants make it extraordinarily easy to generate new code (a developer accepts a suggestion with a single keystroke). But the same tools rarely propose reusing an existing function elsewhere in the codebase, partly because their limited context window prevents them from understanding the full system architecture. The result is a systematic incentive toward duplication and away from the refactoring practices that keep codebases maintainable over time. ",[30,1057,1058],{},"Google's 2024 DORA report corroborates this trade-off, reporting that a 25% increase in AI usage was associated with faster code reviews and better documentation, but also a 7.2% decrease in delivery stability. This surfaces a consistent pattern in that AI accelerates output while potentially eroding the structural qualities that make software maintainable, secure and reliable in the long term. ",[30,1060,1061],{},"For engineering leaders, this creates a paradox. Teams appear more productive in that they are shipping more code, faster. But the total cost of ownership is increasing as duplicated code multiplies maintenance burden, structural inconsistencies make onboarding harder, and the absence of refactoring causes codebases to calcify. The initial speed gains are eventually consumed by the overhead of managing a codebase that was generated quickly but never designed. ",[25,1063,1065],{"id":1064},"why-traditional-testing-practices-are-insufficient","Why Traditional Testing Practices Are Insufficient ",[30,1067,1068],{},"Vibe coding does not just introduce new categories of defect, it undermines the practices that traditionally catch defects before they reach production. ",[30,1070,1071],{},"The most fundamental problem is the comprehension gap. When developers do not read or fully understand the code being generated, they cannot write meaningful tests for it. Effective test design requires understanding of what the code does when given expected inputs, as well as how it should behave at boundaries, under error conditions, and in edge cases. A developer who has described a feature in natural language and accepted the generated implementation without studying it lacks the mental model needed to identify which scenarios require testing. Therefore, test coverage may end up appearing adequate by line count, but is in fact missing the failure modes that matter most. ",[30,1073,1074,1075,1079],{},"There is also an impact on the traditional code review process. CodeRabbit's ",[197,1076,1078],{"href":976,"rel":1077},[201],"data"," shows that AI-generated pull requests create heavier review workloads. This isn’t seen in just more issues per PR, but in wider variance in issue severity, meaning reviewers must spend more time triaging. At the 90th percentile, AI PRs contain 26 issues compared to the human baseline of 12 issues. This volume of review work is difficult to sustain when teams are simultaneously under pressure to ship faster – the same pressure that motivated adopting AI coding tools in the first place. ",[30,1081,1082],{},"This has introduced a speed-quality tension. The traditional cycle of write, review, test, fix was designed for human-paced development. When AI can generate thousands of lines of code in minutes, manual review and testing become bottlenecks, and there can be temptation to relax quality gates rather than slow down delivery. This is precisely how quality debt accumulates. ",[30,1084,1085],{},"There is also an emerging problem with what might be called \"shadow AI development\" – where employees outside formal development teams are building applications and automations using AI coding tools, without engineering oversight or quality governance. These tools have lowered the barrier to creating functional software to the point where non-developers can produce working applications. But \"working\" and \"production-ready\" are very different standards, and organisations are discovering vibe-coded internal tools that lack authentication, contain hardcoded credentials or have no error handling, are deployed and in use before engineering teams are even aware they exist. ",[25,1087,1089],{"id":1088},"what-responsible-ai-assisted-development-looks-like","What Responsible AI-Assisted Development Looks Like ",[30,1091,1092],{},"This isn’t to say that teams should abandon AI coding tools. The productivity benefits are significant, even if they come with caveats. However, what this new landscape requires is a rethinking of how testing and quality practices operate in an AI-assisted development environment. ",[697,1094,1096],{"id":1095},"test-driven-development","Test-driven development  ",[30,1098,1099],{},"Writing tests before AI generates code ensures the code meets specific requirements, regardless of how it was produced. When a developer specifies expected behaviour through tests first, the AI-generated implementation can be validated immediately against those expectations. This approach transforms the developer's role from code reviewer (which the comprehension gap undermines) to specification author (which requires domain knowledge the developer still possesses). TDD also naturally constrains the AI's output – when the generated code must pass predefined tests, many categories of defect are caught at the point of creation rather than downstream. ",[697,1101,1103],{"id":1102},"automated-security-scanning","Automated security scanning  ",[30,1105,1106],{},"CI\u002FCD pipelines must enforce static application security testing (SAST), dynamic application security testing (DAST), and software composition analysis (SCA) on all code, with no distinction between human-written and AI-generated contributions. Given that AI-generated code introduces security vulnerabilities in 45% of cases and contains 2.74 times more XSS vulnerabilities than human code on average, security scanning is not an optional quality enhancement. Organisations should also implement dependency scanning and licence checking, since AI tools can incorporate outdated or insecure third-party libraries without vetting. ",[697,1108,1110],{"id":1109},"property-based-and-contract-testing","Property-based and contract testing  ",[30,1112,1113],{},"Where developers lack the detailed understanding needed to write comprehensive example-based tests, property-based testing offers an alternative: defining the properties that outputs should always satisfy (this function should never return a negative value; this API response should always contain these required fields) and generating test inputs automatically to verify those properties hold. Similarly, contract testing – defining the expected interfaces between services – provides a framework for validating AI-generated code against architectural constraints that the AI itself was not aware of. ",[697,1115,1117],{"id":1116},"code-quality-gates","Code quality gates  ",[30,1119,1120],{},"The GitClear data on code duplication and declining refactoring suggests that quality standards cannot rely on developer discipline alone, they must be embedded in the pipeline. This means automated checks for code duplication thresholds, complexity metrics, test coverage requirements and architectural conformance. AI-generated code that fails these checks should be rejected automatically, just as any code that fails existing CI\u002FCD gates would be. Some organisations are also implementing \"AI code review\" tools – ironically, using AI to review AI-generated code – which adds an additional layer of automated scrutiny. ",[697,1122,1124],{"id":1123},"agentic-security-tools","Agentic security tools  ",[30,1126,1127],{},"Security needs to be embedded in the act of creation rather than added on further downstream. This means security analysis tools that operate as companions to AI coding assistants within the development environment itself, providing real-time feedback on the security implications of generated code as it is produced, rather than catching issues hours or days later in a CI\u002FCD pipeline. This helps to shrink the gap between code generation and security validation near zero. ",[25,1129,1131],{"id":1130},"the-regulatory-dimension","The Regulatory Dimension ",[30,1133,1134],{},"The regulatory environment is adding both urgency and legal liability to AI code quality. ",[30,1136,1137],{},"The EU Cyber Resilience Act (CRA), which came into force in 2024 with compliance deadlines extending through 2027, requires manufacturers of software products to implement comprehensive cybersecurity requirements. Products must be developed according to secure-by-design principles, delivered free from known exploitable vulnerabilities, and supported by ongoing security updates. AI generated software that has never been reviewed by a human with the expertise to assess its security posture is unlikely to meet these obligations. For organisations operating in or selling into the EU market, this creates direct compliance exposure. ",[30,1139,1140],{},"The EU AI Act adds further requirements for software systems that incorporate AI capabilities which increasingly means any software built using AI coding tools, since the generated code itself may embed AI-powered features. High-risk AI systems require comprehensive testing for accuracy, robustness and non-discrimination, with documentation requirements that assume human oversight of the development process. ",[30,1142,1143],{},"In the UK, the Product Security and Telecommunications Infrastructure Act and forthcoming cyber security regulations apply similar principles to connected products and digital services. Sector-specific regulators – the FCA for financial services, the ICO for data protection, the CQC for healthcare – are applying existing regulatory frameworks to software quality in ways that create implicit requirements for code review and testing clarity. ",[30,1145,1146],{},"The implication is that regulatory frameworks consistently place responsibility for software quality on the organisation that deploys it, regardless of whether the code was written by a human developer, generated by an AI tool or produced through vibe coding. This means that engineering leaders who do not establish governance frameworks for AI-generated code are accepting regulatory risk on behalf of their organisations. ",[25,1148,1150],{"id":1149},"theorganisationalpolicy-question","The Organisational Policy Question ",[30,1152,1153],{},"The evidence points clearly toward a need for explicit organisational policy on AI-assisted development. This is not to prohibit it, but to establish the governance framework within which it operates safely. ",[30,1155,1156],{},"At minimum, this means defining where on the spectrum of AI-assisted development the organisation is willing to operate, and under what conditions. For production systems, the expectation should be clear: all code, regardless of origin, must be reviewed, tested and understood by a qualified human before deployment. For prototyping, internal tools and exploratory work, the tolerance for less rigorous oversight may be higher, but even here, security scanning and basic quality gates should apply. ",[30,1158,1159,1160,1164],{},"It also means addressing the skills dimension. The World Quality ",[197,1161,1010],{"href":1162,"rel":1163},"https:\u002F\u002Fwww.capgemini.com\u002Finsights\u002Fresearch-library\u002Fworld-quality-report-2025-26\u002F",[201]," 2025 found that 50% of organisations lack AI\u002FML expertise in their quality engineering teams, which is a gap that extends to understanding the specific failure modes of AI-generated code. Engineers need training not just in using AI coding tools effectively, but in reviewing AI-generated output critically, such as recognising the patterns of duplication, the categories of security vulnerability, and the architectural anti-patterns that these tools systematically produce. ",[30,1166,1167],{},"The most mature organisations are treating AI-generated code as a catalyst for strengthening universal quality practices. If AI-generated code requires comprehensive security scanning, automated quality gates, mandatory test coverage and architectural conformance checks, then so does all code. The AI coding revolution may ultimately leave its most lasting impact not through the code it generates, but through the quality practices it forces organisations to adopt. ",[25,1169,1171],{"id":1170},"the-compounding-problem","The Compounding Problem ",[30,1173,1174],{},"What makes the vibe coding quality crisis particularly challenging is its compounding nature. AI-generated code that is not properly tested accumulates as technical debt. That technical debt makes the codebase harder to understand, which makes it harder to write effective tests, which makes it more likely that the next round of AI-generated code will introduce further undetected issues. GitClear’s 2024 data that states an eightfold increase in code duplication results in a 40% decline in refactoring represents the early stages of this compounding cycle. ",[30,1176,1177],{},"The organisations that act now, establishing testing standards, security gates and governance frameworks for AI-assisted development, will be those that capture the genuine productivity benefits of AI coding tools. However, those that delay, assuming the problem will resolve itself as models improve, are ignoring the clearest finding in the research – code quality does not improve with model size. Better models may produce more syntactically correct code, but as it stands do not result in more secure, maintainable, or architecturally sound code.",{"title":40,"searchDepth":41,"depth":41,"links":1179},[1180,1181,1182,1183,1190,1191,1192],{"id":966,"depth":41,"text":967},{"id":1026,"depth":41,"text":1027},{"id":1064,"depth":41,"text":1065},{"id":1088,"depth":41,"text":1089,"children":1184},[1185,1186,1187,1188,1189],{"id":1095,"depth":912,"text":1096},{"id":1102,"depth":912,"text":1103},{"id":1109,"depth":912,"text":1110},{"id":1116,"depth":912,"text":1117},{"id":1123,"depth":912,"text":1124},{"id":1130,"depth":41,"text":1131},{"id":1149,"depth":41,"text":1150},{"id":1170,"depth":41,"text":1171},"AI coding tools are now embedded in most development workflows, but AI-generated code introduces more security vulnerabilities, duplication and critical defects than human-written code. This article examines the risks and the testing and governance practices engineering leaders need to capture the productivity benefits without accumulating quality debt.","2026-04-29T13:38:42.140Z","\u002Fimg\u002Fblog\u002Fai-engineering-impact-on-testing.png",{},"\u002Fblog\u002Fwhat-ai-assisted-engineering-means-for-software-testing",{"text":1199,"minutes":1200,"time":1201,"words":1202},"12 min read",11.76,705600,2352,{"title":946,"description":954},"blog\u002Fwhat-ai-assisted-engineering-means-for-software-testing","cfRtxM8DxE0UdeHTwyS03eeSUDEdTw9reHcOWIkc8pc",{"id":1207,"title":1208,"author":40,"blogTags":1209,"body":1211,"customExcerpt":1238,"date":1239,"description":1215,"excerpt":47,"extension":44,"h1":47,"image":1240,"meta":1241,"navigation":53,"path":1242,"readingStats":1243,"seo":1248,"stem":1252,"__hash__":1253},"blog\u002Fblog\u002Faudacia-shortlisted-in-british-data-awards-2026.md","Audacia shortlisted in British Data Awards 2026",[1210],"News",{"type":22,"value":1212,"toc":1236},[1213,1216,1224,1227,1230,1233],[30,1214,1215],{},"We’re thrilled to announce that we’ve been named a Finalist for \"AI Company of the Year\" at the British Data Awards 2026.",[30,1217,1218,1223],{},[197,1219,1222],{"href":1220,"rel":1221},"https:\u002F\u002Fpredatech.co.uk\u002Fbritish-data-awards\u002F",[201],"The British Data Awards"," is an annual quest that sets out to discover and celebrate the UK’s data success stories. This year organisations taking part include fresh-faced start-ups, not-for-profits, technology unicorns, government departments, FTSE 100 heavyweights, and everything in between.",[30,1225,1226],{},"Our shortlisting is for a number of successful AI projects delivered across multiple industries, including an AI assistant co-developed with the Medicines and Healthcare products Regulatory Agency (MHRA) that now resolves 80% of complex medicines enquiries within 10 working days; a machine learning model for STERIS that predicts sterilisation dosage with 97%+ accuracy; and an AI-powered WhatsApp chatbot for Northern Trains that handles over one million customer conversations a year, cutting voice call volumes by 25%.",[30,1228,1229],{},"For it's sixth edition the British Data Awards received a record number of entries at 462, so we’re especially pleased to be announced as a Finalist.",[30,1231,1232],{},"As always, this recognition is a reflection of the amazing hard work of our brilliant teams and wonderful client partnerships.",[30,1234,1235],{},"Congratulations to all of the other finalists!",{"title":40,"searchDepth":41,"depth":41,"links":1237},[],"Audacia is shortlisted for AI Company of the Year at the 2026 British Data Awards. ","2026-03-27T14:51:16.422Z","\u002Fimg\u002Fblog\u002Fbritishdataawards2026-shortlisted.png",{},"\u002Fblog\u002Faudacia-shortlisted-in-british-data-awards-2026",{"text":1244,"minutes":1245,"time":1246,"words":1247},"2 min read",1.345,80700,269,{"title":1249,"description":1250,"keywords":1251},"Audacia shortlisted in British Data Awards 2026 ","AI consultancy, Audacia, is shortlisted for AI Company of the Year at the 2026 British Data Awards. ","AI company of the year, British Data Awards, award-winning consultancy, AI consultancy","blog\u002Faudacia-shortlisted-in-british-data-awards-2026","OBySSEsi-DDdNsSpGnksIuxyiYELGKwDweD8lXb8XHs",{"id":1255,"title":1256,"body":1257,"companyName":1261,"content":1262,"description":40,"extension":44,"largeCoverImage":1263,"meta":1264,"navigation":53,"path":1265,"seo":1266,"stem":1267,"__hash__":1268},"testimonials\u002Ftestimonials\u002Ftom-broadbent-aesseal-plc.md","Tom Broadbent",{"type":22,"value":1258,"toc":1259},[],{"title":40,"searchDepth":41,"depth":41,"links":1260},[],"AESSEAL plc","The way that we work is that we are subject matter experts, we know our business, we know our customers, we can then have that conversation with the team at Audacia. It is very much a collaborative 2 way process and the level of communication is just fantastic.","\u002Fimg\u002Faesseal-cover-image.jpg",{},"\u002Ftestimonials\u002Ftom-broadbent-aesseal-plc",{"title":1256,"description":40},"testimonials\u002Ftom-broadbent-aesseal-plc","l3_mVF7K3qSl27RqOciuSVhu3snnYdbe78vpkN6rHRI",{"id":1270,"title":1271,"body":1272,"description":40,"extension":44,"groupedServices":1276,"h1":1393,"h2":1394,"meta":1395,"navigation":53,"path":1401,"projects":1402,"seo":1404,"stem":1414,"__hash__":1415},"servicePage\u002Fservice-page.md","Digital Transformation & Software Development Services | Audacia",{"type":22,"value":1273,"toc":1274},[],{"title":40,"searchDepth":41,"depth":41,"links":1275},[],[1277],{"title":1278,"groups":1279},"Services",[1280,1308,1336,1364],{"title":329,"subtitle":1281,"services":1282},"Building large-scale, complex software platforms - from enterprise legacy modernisation, to greenfield projects, we deliver solutions that automate, streamline and simplify processes, meeting the needs of today and the future. ",[1283,1288,1293,1298,1303],{"slug":1284,"title":1285,"subtitle":1286,"image":1287},"software-development-company-uk","Software Development","Engineering complex, mission critical software platforms","\u002Fimg\u002Fservice-page\u002Fsoftware-development.jpg",{"slug":1289,"title":1290,"subtitle":1291,"image":1292},"software-testing-company","Software Testing","Ensuring systems are performant, accessible, scalable and secure","\u002Fimg\u002Fservice-page\u002Fsoftware-testing.jpg",{"slug":1294,"title":1295,"subtitle":1296,"image":1297},"systems-integration","Systems Integration","Integrating systems to operate as a single framework","\u002Fimg\u002Fservice-page\u002Fsystem-intergration.jpg",{"slug":1299,"title":1300,"subtitle":1301,"image":1302},"mobile-development","Mobile Development","Delivering performant, scalable and robust user-driven mobile applications","\u002Fimg\u002Fservice-page\u002Fmobile-development.jpg",{"slug":1304,"title":1305,"subtitle":1306,"image":1307},"front-end-development-company","Front-end Development","Building intuitive and scalable user-focused digital experiences","\u002Fimg\u002Fservice-page\u002Ffront-end-development.jpg",{"title":628,"subtitle":1309,"services":1310},"Turning data into a valuable asset - from building centralised data platforms to delivering clear, accessible reporting and analytics, we enable teams to make faster, more informed decisions based on trusted data.",[1311,1316,1321,1326,1331],{"slug":1312,"title":1313,"subtitle":1314,"image":1315},"data-visualisation-consultancy","Data Visualisation & Reporting","Delivering custom reporting for data-driven decision making","\u002Fimg\u002Fservice-page\u002Freporting.jpg",{"slug":1317,"title":1318,"subtitle":1319,"image":1320},"data-analytics-company","Data Analytics","Uncovering actionable insights for data-driven transformation","\u002Fimg\u002Fservice-page\u002Fanalytics.jpg",{"slug":1322,"title":1323,"subtitle":1324,"image":1325},"data-migration-services","Data Migration","Delivering large migrations across architectures and environments","\u002Fimg\u002Fservice-page\u002Fmigration.jpg",{"slug":1327,"title":1328,"subtitle":1329,"image":1330},"data-engineering-services","Data Engineering","Engineering pipelines to consolidate, process and aggregate data","\u002Fimg\u002Fservice-page\u002Fdata-engineering.jpg",{"slug":1332,"title":1333,"subtitle":1334,"image":1335},"data-platform-consulting","Enterprise Data Products","Designing, delivering and optimising enterprise data platforms to provide unified data foundations","\u002Fimg\u002Fservice-page\u002Fenterprise-data-products.jpg",{"title":627,"subtitle":1337,"services":1338},"Guiding organisations through the practical application of AI - from custom machine learning models to generative AI, we create solutions that improve services and enable organisations to make decisions better, faster and at scale.",[1339,1344,1349,1354,1359],{"slug":1340,"title":1341,"subtitle":1342,"image":1343},"ai-consultancy","Research & Prototyping","Validating AI concepts through research and prototyping","\u002Fimg\u002Fservice-page\u002Fprototype.jpg",{"slug":1345,"title":1346,"subtitle":1347,"image":1348},"generative-ai-company","Generative AI","Building custom GenAI solutions across image, vision and text","\u002Fimg\u002Fservice-page\u002Fgen-ai.jpg",{"slug":1350,"title":1351,"subtitle":1352,"image":1353},"ai-machine-learning-company","Concept to Delivery","Taking AI and machine learning solutions from PoC to production","\u002Fimg\u002Fservice-page\u002Fconcept-to-delivery.jpg",{"slug":1355,"title":1356,"subtitle":1357,"image":1358},"ai-machine-learning-operations-consulting","AI Cloud Services & MLOps","Providing machine learning lifecycle management to build, deploy and scale trusted AI solutions rapidly","\u002Fimg\u002Fservice-page\u002Fai-cloud.jpg",{"slug":1360,"title":1361,"subtitle":1362,"image":1363},"ai-machine-learning-consulting","Machine Learning","Building custom machine learning models optimised to specific challenges, data and operational environments","\u002Fimg\u002Fservice-page\u002Fml.jpg",{"title":1365,"subtitle":1366,"services":1367},"Cloud","Supporting organisations in building secure, scalable cloud environments - whether migrating systems, modernising platforms, or optimising existing infrastructure, we focus on delivering cloud solutions that improve performance, resilience, and cost-efficiency.",[1368,1373,1378,1383,1388],{"slug":1369,"title":1370,"subtitle":1371,"image":1372},"cloud-architecture-consulting","Cloud Architecture & Design","Designing cloud solutions optimised for resilience and scale","\u002Fimg\u002Fservice-page\u002Fcloud-design.jpg",{"slug":1374,"title":1375,"subtitle":1376,"image":1377},"cloud-migration-consulting","Cloud Adoption & Migration","Migrating to cloud infrastructures to enhance accessibility, performance and scalability","\u002Fimg\u002Fservice-page\u002Fcloud-adoption.jpg",{"slug":1379,"title":1380,"subtitle":1381,"image":1382},"cloud-application-development","Cloud Native Solutions","Building scalable and resilient cloud native solutions rapidly","\u002Fimg\u002Fservice-page\u002Fcloud-native.jpg",{"slug":1384,"title":1385,"subtitle":1386,"image":1387},"devops-consulting-services","DevOps & Pipelines","Embedding automation through DevOps for continuous integration, delivery and deployment","\u002Fimg\u002Fservice-page\u002Fdevops.jpg",{"slug":1389,"title":1390,"subtitle":1391,"image":1392},"cloud-security-consulting","Cloud Security","Securing cloud environments through architecture, automation and optimisation","\u002Fimg\u002Fservice-page\u002Fcloud-security.jpg","Your end-to-end\u003Cbr>technology delivery partner","We're trusted to deliver large-scale, complex technology projects for leading organisations across engineering, data, AI and cloud.",{"serviceBlock":1396,"projectsTitle":52,"testimonial":1400},{"title":47,"description":47,"services":1397},[1284,1304,1398,1350,1327,1294,1399,1289],"mobile-app-development-company","cloud-migration-services","steve-evans-specialist-marine-consultants-smc","\u002Fservice-page",[56,57,58,1403],"aesseal-quotation-platform",{"title":1271,"keywords":1405,"description":1406,"jsonSchema":1407},"Software development company, Software development, Collaborative development, Software consulting, Web applications, Mobile applications, Systems integration, Digital transformation, Digital transformation company, Digtal transformation partner","Digital transformation & software development company, Audacia, specialise in helping leading organisations improve, scale and innovate through technology. ",[1408,1410,1412],{"json":1409},"{\n  \"@context\": \"https:\u002F\u002Fschema.org\",\n  \"@type\": \"Corporation\",\n  \"name\": \"Audacia - Software Development Company\",\n  \"alternateName\": \"Audacia - Software Development Company\",\n  \"url\": \"https:\u002F\u002Faudacia.co.uk\u002F\",\n  \"logo\": \"https:\u002F\u002Faudacia.co.uk\u002Flogo.svg\",\n  \"contactPoint\": [{\n    \"@type\": \"ContactPoint\",\n    \"telephone\": \"0113 543 1300\",\n    \"contactType\": \"sales\",\n    \"areaServed\": \"GB\",\n    \"availableLanguage\": \"en\"\n  },{\n    \"@type\": \"ContactPoint\",\n    \"telephone\": \"0203 855 6040\",\n    \"contactType\": \"sales\",\n    \"areaServed\": \"GB\",\n    \"availableLanguage\": \"en\"\n  }],\n  \"sameAs\": [\n    \"https:\u002F\u002Fwww.linkedin.com\u002Fcompany\u002Faudacia-consulting\u002F\",\n    \"https:\u002F\u002Ftwitter.com\u002Faudacia_tech\",\n    \"https:\u002F\u002Fwww.facebook.com\u002FAudaciaTechnology\"\n  ]\n}",{"json":1411},"{\n  \"@context\": \"https:\u002F\u002Fschema.org\",\n  \"@type\": \"ProfessionalService\",\n  \"name\": \"Audacia - Software Development Company\",\n  \"image\": \"https:\u002F\u002Faudacia.co.uk\u002Flogo.svg\",\n  \"@id\": \"\",\n  \"url\": \"https:\u002F\u002Faudacia.co.uk\u002F\",\n  \"telephone\": \"0113 543 1300\",\n  \"address\": {\n    \"@type\": \"PostalAddress\",\n    \"streetAddress\": \"46 The Calls\",\n    \"addressLocality\": \"Leeds\",\n    \"postalCode\": \"LS2 7EY\",\n    \"addressCountry\": \"GB\"\n  },\n  \"geo\": {\n    \"@type\": \"GeoCoordinates\",\n    \"latitude\": 53.794709,\n    \"longitude\": -1.537110\n  },\n  \"openingHoursSpecification\": {\n    \"@type\": \"OpeningHoursSpecification\",\n    \"dayOfWeek\": [\n      \"Monday\",\n      \"Tuesday\",\n      \"Wednesday\",\n      \"Thursday\",\n      \"Friday\"\n    ],\n    \"opens\": \"09:00\",\n    \"closes\": \"17:30\"\n  },\n  \"sameAs\": [\n    \"https:\u002F\u002Fwww.linkedin.com\u002Fcompany\u002Faudacia-consulting\u002F\",\n    \"https:\u002F\u002Ftwitter.com\u002Faudacia_tech\",\n    \"https:\u002F\u002Fwww.facebook.com\u002FAudaciaTechnology\",\n    \"https:\u002F\u002Faudacia.co.uk\u002F\"\n  ] \n}",{"json":1413},"{\n  \"@context\": \"https:\u002F\u002Fschema.org\u002F\", \n  \"@type\": \"BreadcrumbList\", \n  \"itemListElement\": [{\n    \"@type\": \"ListItem\", \n    \"position\": 1, \n    \"name\": \"Home\",\n    \"item\": \"https:\u002F\u002Faudacia.co.uk\u002F\"  \n  },{\n    \"@type\": \"ListItem\", \n    \"position\": 2, \n    \"name\": \"Services\",\n    \"item\": \"https:\u002F\u002Faudacia.co.uk\u002Fservices\"  \n  }]\n}","service-page","guI5P__UY2w9K1r33oL5elVC5aiqOr0zrOLg49LJrw8",1778145929771]