MetaNotes is an AI meeting assistant that joins Google Meet, Teams, and Zoom calls and turns them into transcripts, summaries, and conversation analytics after the meeting. It's a public product, and internally it also replaced a paid third-party notetaker. I owned the frontend end to end.
What I built
- Transcript view: Built the frontend that renders a meeting's transcript once it's processed, keeping it fast and navigable even for long, hour-plus calls.
- The dashboard: Designed a dense-but-usable dashboard with one-click AI summaries, sentiment analytics, and topic tagging, so users can get value out of a full transcript without reading the whole thing.
- Frontend ownership: Was the sole frontend engineer from start to production — set up the repo and architecture, made the state-management and structure calls myself, and worked directly with the backend team to turn their AI orchestration into a clean, dependable UI.
Why it mattered
MetaNotes ships as a public product and, used internally, let the company drop its paid transcription vendor — saving around $24,000 a year. Either way it takes manual note-taking off people's plates and hands back summaries and insights right after a call.