Inspiration Web accessibility is often treated as a checklist rather than a real user need. Many tools give scores but don’t help developers fix issues. We were inspired to build something that bridges the gap between detection and action—making accessibility understandable, actionable, and fast to improve for real-world teams.

What it does

Lumio is an AI-powered accessibility scanner. You input a website URL, and it analyzes pages to detect accessibility issues. It explains each issue in simple language, shows who is affected, highlights the exact elements, and provides code-level fixes. It also groups repeated issues, prioritizes critical problems, tracks regressions, and enables report exports.

How we built it

We combined automated accessibility auditing with AI-based explanation and summarization. A scanning engine crawls pages and detects issues, while an AI layer converts technical findings into plain English insights and actionable fixes. We also built logic to cluster issues, map them to code (when repos are available), and track changes across scans.

Challenges we ran into

One major challenge was translating complex accessibility rules into clear, useful explanations. Another was accurately mapping issues back to specific code elements across different frameworks. Handling large websites efficiently and reducing duplicate issue noise also required careful optimization.

Accomplishments that we're proud of

We created a tool that doesn’t just report problems but helps solve them. The ability to group issues, prioritize fixes, and provide developer-ready solutions makes Lumio practical and impactful. Turning accessibility into something approachable is a key achievement.

What we learned

We learned that developers value clarity over scores. Simplifying technical concepts without losing accuracy is difficult but essential. We also saw how automation plus AI can significantly reduce time spent on debugging accessibility issues.

What's next for Lumio

We plan to enhance real-time scanning, improve code mapping accuracy, and integrate directly into developer workflows (CI/CD, IDEs). We also aim to expand support for dynamic content and provide deeper insights tailored to different user groups.

Built With

Share this project:

Updates