Harsh Gurawaliya Back to Projects

Agentic VC Investment Analyst

Startup Squirrel

The agentic VC analyst that hoards startup signals and cracks open investment theses.

How It Works Data Sources

Demo: Startup Squirrel pulls evidence, debates the case, critiques itself, and scores the investment thesis.

The essential pipeline

A structured workflow keeps the recommendation grounded and auditable.

1

Decompose

Breaks the company into team, traction, market, moat, momentum, and risk.

2

Collect Evidence

Pulls cited signals from the connected sources.

3

Generate

Builds pro and con arguments tied to evidence.

4

Critique

A devil's-advocate pass stress-tests the arguments.

5

Refine

Weak claims are downgraded after critique.

6

Score

Outputs a 0-100 score and confidence level.

7

Decide

Returns Invest, Hold, or Pass with rationale.

Connected signals

Eight data sources feed the analysis

The tool connects to News, GitHub, Hacker News, Reddit, Hiring, Web traffic, Funding, and Trends before it generates pro and con arguments.

  • Market, builder, hiring, traction, and funding signals in one place.
  • Every claim can be traced back to the supporting source.
Startup Squirrel connected data sources: News, GitHub, Hacker News, Reddit, Hiring, Web traffic, Funding, and Trends
The connected source panel used by Startup Squirrel.

Example input

Quantum Systems thesis run

A sample robotics thesis run for Quantum Systems, including startup context, industry, description, and investment-relevant notes before analysis starts.

Startup Squirrel Quantum Systems example input form with company, industry, description, context, and Crack the thesis button
The thesis input screen: company, industry, what it does, context, and the action that starts the multi-source analysis.
Startup Squirrel Quantum Systems source signals showing Funding, GitHub, News, Reddit, Hiring, Hacker News, Web traffic, and Trends cards
Evidence view: 9 signals across funding, developer traction, press, community sentiment, hiring, web traffic, and trends.
Startup Squirrel Quantum Systems recommendation screenshot showing Lean invest, 56 composite score, 62 percent confidence, pro score 68, and con score 56
Final output: Lean invest, 56/100 composite score, 62% confidence, and pro/con arguments with cited evidence.

Pluggable Investment Strategies

Different funds can score the same company with different thesis profiles.

Why this matters

A pre-seed SaaS fund, deep-tech fund, growth fund, and contrarian fund should not use the same scoring logic. Startup Squirrel keeps the analysis engine reusable while making the fund thesis configurable.