AI agent for venture capitalists
Startup Signal Sourcing Agent
A weekly short list of new companies matching the thesis, free of known names, that improves as the partner gives feedback
What it does
Most deals come through the investor's own network, so good companies outside it are missed. Signals such as a burst of hiring, a product launch or growing open-source activity appear weeks before a funding round. This agent watches those public signals: job posts, launch announcements, repository stars and commits, and conference talks. It scores each company against the fund's thesis, such as stage, sector and geography. Then it checks the CRM for existing relationships and prior passes, and drops those already seen. It reviews which hits the partner kept and which were dismissed and refines its keywords and weights, then scans again. The partner approves any outreach. Edge case: a company that was passed on 18 months ago has doubled its team, and the agent shows the earlier note.
How it works
Follow the arrows from top to bottom. The orange dashed arrow is the loop: when a check fails, the agent goes back and tries again.
Read the steps as a list
- Weekly scan begins
- Search job posts, launches and repository activity for thesis keywords
- Score each company for thesis fit and signal strength
- Check the CRM for relationships and earlier passes
- Is the company new, or has something material changed since the pass?If not: drop it from the list and keep a note. Back to step 2.
- Write a short profile with the signals and why it fits
- Partner reviews the list and approves outreachThe agent waits here for your OK.
- Record which companies were kept or dismissed and why
- Did at least 3 in 10 shortlisted companies pass the partner's review?If not: refine keywords and weights from the dismissed reasons and rescan. Back to step 2.
- Updated list and revised keywords saved
How it decides
It scores companies on thesis fit and signal strength, excludes known companies unless something material changed, and adjusts weights from the partner's keep and drop choices.
- Keep companies scoring 70 or higher out of 100
- Re-surface a passed company when headcount rises 50% or it raises funding
- Drop companies outside the fund's geography
- Adjust keyword weights after each review
Make it yours
Every agent is a starting point. You choose these settings for your own situation.
- Thesis keywords
- Score cut-off (default 70)
- Geography filter
- Scan day
What keeps you in control
It always asks you first
- Any outreach to a founder
Hard limits
- Never contacts founders
- Uses public information only
- Records the source of each signal
It stops when
- Done: weekly list delivered
- Stop: CRM unavailable so known names cannot be excluded
Set it up
We guide you through the set-up, step by step
Members get the full set-up guide for this agent. No technical skills needed: you copy, paste and upload.
- One set of instructions to paste into your AI, with the clicks for ChatGPT, Claude, Microsoft 365 Copilot, Gemini and Grok
- The agent then walks you through connecting your own data, one source at a time
- A downloadable copy with the flow chart, the rules and the full guide
