Let TimeCampus name the project behind each captured hour

Turning captured activity into billable time means knowing which project each hour belongs to. TimeCampus is being built to make that attribution honest and transparent.

The platform is designed to score each captured event against your project names using a deterministic keyword classifier first, then optionally refine the match with an LLM. Either way, the result is stored as a suggestion — labelled with the method and model that produced it — and the suggested project would never become the real project until you explicitly accept it.

The in-timeline accept and reject controls are the next piece being built, so this describes the attribution experience TimeCampus is working toward rather than a completed UI flow today.