Accessibility in Power BI is widely recognised as important, but it’s rarely approached in a consistent or scalable way. Most teams rely on manual checks and visual judgement, making it difficult to know whether a report is truly usable for everyone.

In this session, I’ll share how I built PBIX A11y, a free, browser-based accessibility checker for Power BI, using AI to turn a recurring frustration into a practical solution.

This is a behind-the-scenes look at the journey, from identifying the gap and reframing the problem from a user perspective, to designing meaningful checks and making deliberate decisions around usability, trust, and compliance. I’ll also share how Copilot became a prompting companion throughout the process, helping refine ideas, iterate faster, and expose the limitations that required better prompting and human judgement. I’ll show how AI accelerated development, but also where domain knowledge was essential to ensure outputs were accurate, useful, and responsible.

Beyond the technical build, I’ll cover real-world considerations that emerged along the way, including validating AI-generated results, navigating GDPR and privacy requirements, SEO trade-offs for a public tool, and the constraints these factors introduce.

Attendees will leave with a practical perspective on what it really takes to turn an idea into a purposeful AI solution, including the trade-offs and constraints involved.