RESEARCH COLLABORATION · CONA LAB
AI Tool / Accessibility auditing
SiteSense
An accessibility auditor that flags WCAG-ordered issues in under 30 seconds, built with the CoNA Lab.
<30s · Audit time

- OWNERSHIP
- Research collaboration
- CLIENT
- CoNA Lab
- OUR ROLE
- UX design and interface engineering, with the lab's backend developer
- STATUS
- Working prototype
Prototype, not a deployed service. We can walk through the audit flow and the WCAG mapping on request.
01 · Problem
Manual audits couldn't keep pace with the state's web footprint.
Maryland state web properties had to improve accessibility, but the standard process — manual review or an off-the-shelf scanner — was slow and hard to scale across hundreds of pages. Auditors spent their time finding issues instead of fixing them.



02 · Idea
Automated auditing, built on vision models and WCAG 2.
The CoNA Lab set out to build Site Sense: an automated auditor using vision models and the lab's own custom API. I partnered with the backend developer and owned the UX design and interface engineering, structuring audit results around WCAG 2 so an auditor could act on them in order of severity rather than scroll through a raw list.
03 · Result
A prototype faster than the industry incumbents.
After weeks of collaboration we shipped a testable prototype ready for real auditors. Typical tools took minutes to scan a site; Site Sense returned full results in under 30 seconds, with issues grouped by WCAG criterion and severity.


<30s
AUDIT TIME
Prototype timing during CoNA Lab reviews · same pages compared with two incumbent tools
HOW THESE NUMBERS WERE MEASURED
Audit time was measured on the working prototype during CoNA Lab reviews, timing the same pages that two incumbent tools had already scanned. It is a single-team measurement, not a published benchmark, and the prototype is not deployed publicly.
Next project
Tracky

Studio Product / React analytics SDK