Finding new companies

Once a day the system reads companies announcing their own funding rounds, confirms each company on its own site, and adds it to the shared archive with the round as a dated, linked event. Your watchlist draws on that archive, so a company with a fresh round can reach it with the round as the reason.

The radar can only watch companies it knows about. Without this it knew the ones somebody typed in, and it never grew.

It reads what companies say about themselves, not what is reported about them. Most news publishers forbid commercial use of their content or block AI crawlers outright, and a company's own announcement is better evidence anyway. Every source carries the terms decision that admitted it.

It is deliberately strict. A seed round, a fund, debt, an announcement on somebody else's site, or a domain the company's own homepage does not confirm are all skipped. A wrong company in the archive costs more than a missing one, because everything downstream is keyed on its domain.

1. 01It is off until an operator sets COMPANY_DISCOVERY=on. From then it runs daily with nothing else to set up.
2. 02Companies it adds appear in the archive with a funding event you can open.
3. 03Check the admin screen for each source’s last run and any error. Every source can be switched off on its own.

• Signals and collectors
• Companies to watch
• Grounding and evidence