TAMSIV Has a Face — First Demo Video + Quality Sprint

For 6 months, TAMSIV existed only as text and screenshots. Today, for the first time, the app has a face: a full demo video. Why a video now? A voice app needs to be seen . Reading "speak, AI understands and organizes" isn't enough — you need to see the mic activating, the real-time transcription, the task being created automatically in the right folder. The video shows all of that in 2 minutes. It's now embedded directly in the website's hero section . Quality sprint: 7 commits, 0 new features Alongside the video, I spent a week on a 100% quality sprint. No new features — only polish, reliability, and silent fixes. Dictaphone: 100% reliable stop button The most frustrating bug: sometimes the "stop" button wouldn't respond. The mic kept recording, forcing users to kill the app. The cause? A timing issue between native STT initialization and React state. The fix: A "standby" mode for native STT — ready instantly without blocking the UI Recording start 2x faster (no callback wait) A stop