The shape of your day, in one view.
contextbliss watches your webcam and screens every 30 seconds, classifies what you’re actually doing, and speaks up the moment you switch context.
What it does
Detects your context
One webcam frame plus one frame per shared screen, every 30 seconds. A multimodal model picks the state and the work context — no manual tagging.
Speaks up on shifts
The moment you go from coding to comms, or research to a Zoom, you hear a short spoken nudge. Awareness without checking a screen.
Replays your day
The dashboard leads with one synthesized headline (“Mostly coding — 64% of the past 24 hours”) and breaks down state, context, shifts, and free windows.
How it works
Allow camera & share screens
Browser-native, in your tab. Multi-monitor aware — we ask for each screen so the work signal is real, not just presence at the desk.
Click start
Frames are captured locally, composited, and sent to a multimodal model. Classification returns in seconds; observations stream into a SQLite log on a Fly volume.
Hear it, see it
Context shifts trigger a spoken announcement. The report view shows the timeline, the dominant context, and the shifts — ready when you want to look back.
Ready to see your day?
Sign in with Google, allow camera and screens, and click start. That’s the whole setup.
Open contextbliss →