5 Essential Steps to Build Your Own AI Success
â–Ľ Summary
– Most AI projects fail to progress or deliver results, making a strong foundational strategy essential for success.
– A successful AI initiative must start with robust data governance, good practice, and clear data ownership.
– The focus should be on delivering measurable value to end-users by automating simpler tasks to free up human talent for complex work.
– IT teams must be integrated with the business, acting as partners to drive organizational change and transformation.
– Building a learning culture through iterative, agile projects and effective storytelling is crucial to building confidence and spreading benefits.
Transforming a promising AI concept into a functional, valuable system remains a significant hurdle for many organizations. Studies indicate a majority of AI initiatives stall early on, and even those that advance often fail to deliver substantial returns. Paul Neville, Director of Digital, Data, and Technology at The Pensions Regulator (TPR) in the UK, has focused on converting AI into a tangible competitive edge. He outlines five critical steps for building a successful AI project.
The first step involves establishing a solid technological and data foundation. Neville emphasizes that exploring AI effectively requires robust underlying infrastructure. His team launched a comprehensive Digital, Data, and Technology Strategy, which includes a dedicated data plan to foster adoption of new technologies and standards. Beyond cybersecurity, this work prioritizes strong data governance and clear data ownership. “If your data isn’t in good order, that’s a fundamental problem,” Neville notes. “While AI tools can assist, true strength comes from good practice and governance.”
Success with AI is directly tied to delivering clear value for end-users. The focus must remain on the people within the organization, not the technology itself. “User focus is absolutely critical,” Neville states. “You’re here to deliver value, so find it, measure it, and talk about it.” A practical example at TPR was replacing disparate case management systems with an integrated platform like Microsoft Dynamics 365. This move showcased the benefits of automation, freeing talented staff to handle complex, outlier cases. Automating simpler tasks allows human expertise to be directed where it adds the most value.
For AI to be effective, technology teams must be fully integrated with the broader business. The outdated model of IT as a separate service unit is no longer viable. “Now, IT teams are part of the business,” Neville explains. As a member of the executive team, he ensures other leaders understand how digital and data capabilities can drive internal change and industry-wide transformation. Achieving integration with strong business sponsorship creates a unified team that can make a real difference.
Developing a learning culture is essential for sustained progress. Identifying an AI opportunity is just the beginning. Neville advocates for an iterative, agile approach where teams continuously measure and improve. This product-centered mindset helps build reusable capabilities and spreads new working methods beyond the IT department. An iterative process builds organizational confidence; colleagues begin to believe in the team’s ability to drive meaningful change.
The final component is mastering the art of storytelling. How you communicate about AI is vital. Neville describes this as a major part of his role. To foster a cohesive narrative, TPR established the Pensions Data and Digital Working Group, bringing together trustees, actuaries, lawyers, and technologists from across the sector. The goal is to collaboratively explore how emerging technologies can solve persistent challenges. Instead of imposing mandates, the regulator aims to incentivize innovation through collaboration.
TPR is actively pursuing new AI applications. One use case involves using AI to monitor news sources and link relevant information to specific pension schemes, helping teams identify potential risks they might otherwise miss. Another project leverages generative AI to analyze the climate-related statements companies are required to submit, transforming a manually intensive review process into actionable analysis.
Looking ahead, Neville envisions a more streamlined, data-enabled organization. Internal systems will feel cohesive, reducing manual work and allowing staff to concentrate on high-value activities. This seamless experience will extend to external partners as well. “We’ll be much easier to work with,” he says, “because if you’ve told us something, we’ll know it.”
(Source: ZDNET)
