At an event-management SaaS company, I built an MCP server inside the live product so non-engineers can ask questions in plain English about data, APIs, and business rules. It answers using the real system, is read-only by default, and has end-to-end tests.
I designed an AI-assisted delivery workflow that turns tickets into draft code for human review, runs browser tests automatically, and posts results to Slack. Error alerts also start an agent that drafts a fix.
I lead frontend work on a platform used during high-traffic live events, building real-time features and AI-driven personalization with React and Ionic. I also work directly with enterprise clients to shape requirements and meet event deadlines.
As a freelance engineer, I took over a stalled health-tech mobile app, fixed data-sync bugs, and made it work offline. I also design coding tasks for AI benchmarks, write automatic checkers to grade AI work, and close loopholes that let it cheat.
