AI Scout Agent — Automated Startup Discovery & Scoring
n8n · Multi-Source Data Enrichment · Automated Scoring & Alerting
01The challenge
VC teams and scouts rely on manual research to find promising early-stage startups. It is slow, it does not scale across sources, and it is easy to miss signals entirely when the interesting activity is spread across a dozen platforms that nobody has time to check daily.
02The data problem
The signals that matter live in completely different places and formats: launch activity on Product Hunt, founder chatter on X, engineering traction on GitHub, cohort announcements from YC Demo Days, and long-form context buried in podcasts. None of it shares a common company identifier, and none of it arrives enriched with the funding or founder background needed to judge whether a startup is actually worth a conversation.
03My approach
I built a Scout Agent in n8n that pulls from X, Product Hunt, GitHub, YC Demo Days and podcasts, then enriches each discovered profile against Crunchbase, LinkedIn and AngelList so every candidate arrives with funding activity, founder background and engagement metrics attached rather than as a bare company name.
On top of the enriched profile I built a custom scoring system that flags high-potential startups against weighted signals, so the output is a ranked shortlist rather than another undifferentiated feed. Anything crossing the threshold automatically triggers Notion logging for the permanent record and a Slack alert so the team sees it the same day.
The whole workflow is modular: each data source is built as its own sub-workflow, so a source can be added, fixed or swapped out without touching the scoring and alerting logic downstream.
04The result
The scouting process runs continuously without anyone manually checking sources. High-potential startups surface into Slack with their funding, founder and engagement context already attached, and every candidate is logged in Notion automatically — turning an ad-hoc manual research task into a repeatable, always-on pipeline.
Tools & tech stack
- n8n: orchestration of the full discovery, enrichment, scoring and alerting workflow
- Source integrations: X, Product Hunt, GitHub, YC Demo Days and podcast feeds, each as an isolated sub-workflow
- Enrichment APIs: Crunchbase, LinkedIn and AngelList for funding, founder and company data
- Custom scoring logic: weighted flags across funding activity, founder background and engagement metrics
- Notion: automated logging of every scored candidate
- Slack: real-time alerting for high-potential matches
Have data that isn't telling you anything yet?
Send me what you are working with — even if it is four spreadsheets and a CRM export that don't agree with each other.