Citesvue

Citesvue is what I am heads-down on right now. It is the evidence layer for recorded work. Every team records more than ever. Almost no team turns those recordings into anything that survives the week. Citesvue is what does.

Most meeting AIs summarise. They generate confident-sounding bullets, with no evidence and no visual context. The moment someone asks "where did they actually say that?", there is nothing to point to. That is fine for a quick recap. It falls apart the second the recording is a UAT walkthrough, a regulatory call, or any moment where a wrong claim becomes a real problem.

Citesvue is a different category. It fuses speech, screen, and timing into one queryable layer that lives underneath every recording your team makes. Word-level transcripts. OCR-aware screen analysis. Pronoun resolution. Pointing context. The output is not a summary - it is a cited answer with a quote, a timestamp, and the exact frame it came from.

I built it for the people who get audited. UAT teams, customer-success teams, regulated meetings, and any process where being wrong costs more than being slow. There is a free tier, paid plans that scale with the value, and a security posture built for serious work - TLS in transit, AES at rest, role-based access, full audit log, and raw media deleted after processing.