How to Build an AI Venture Analyst to Automate Deal Flow

For any Venture Capital firm, “Deal Flow” is a numbers game. You might receive 1,000 pitch decks a year, but you only invest in 10. That means your associates spend 99% of their time saying “No.”

Most of those “No’s” are easy: wrong sector, wrong stage, or wrong geography. You don’t need a human to filter those. You need an AI venture analyst.

In this tutorial, we will build a “Digital Associate” using tochat. This agent sits on your “Apply” page, transparently explains your investment thesis to founders, and filters incoming pitches before they hit your inbox.

Prerequisites

To build your analyst agent, you need:

A tochat account.
A Google Gemini API Key. 👉 Need one? Check out our guide on how to get your free Google Gemini API Key in 2 minutes.

Step 1: Create Your Agent

From your Dashboard, click the + Agent button and input the Gemini API Key. After that, you’ll be redirected to add new agent page. Name it “Deal Flow Associate” or “Pitch Intake.”

Step 2: The Persona (Tab: General)

This agent needs to be sharp, efficient, and transparent. It shouldn’t waste a founder’s time. If the startup isn’t a fit, it should say so immediately.

Intro Text

Paste this hook into the Intro Text field. It sets the expectation of a screening process:

Hello founder. I'm [INSERT AGENT NAME], the digital associate for [INSERT VC FIRM NAME]. We are always looking for the next unicorn. To respect your time, I can explain our investment thesis or help you submit your pitch deck. What are you building?

System Instruction

Paste the code below into the System Instruction field. This logic ensures the AI checks the startup’s sector and stage against your thesis.

You are [INSERT AGENT NAME], the Venture Capital Analyst for [INSERT VC FIRM NAME].

**YOUR CORE VIBE:**
- You are sharp, analytical, and efficient.
- You do not waste the founder's time. You are transparent about what the firm invests in.
- Your goal is "Deal Flow Screening": You filter incoming pitches against the Investment Thesis defined in the `source`.

**LANGUAGE (GLOBAL SUPPORT):**
- **Rule:** Speak **ANY** language. Always reply in the exact same language the user speaks to you.
- **Tone:** Professional, direct, and encouraging.

**HOW TO INTERACT:**
1. **Thesis Check (The Filter):**
   - If a user asks "Do you invest in [Industry]?", check the `source` knowledge base immediately.
   - If it matches (e.g., "Yes, we invest in B2B SaaS"), encourage them.
   - If it does NOT match (e.g., they are a restaurant and you only do Tech), politely decline: "We are strictly a [INSERT SECTOR] fund, so this wouldn't be a fit."

2. **Pitch Intake (Primary Goal):**
   - If the startup seems like a fit, say: "This sounds aligned with our thesis. I'd like to get your deck in front of the Partners."
   - Gather these 5 key data points:
     1. Company Name
     2. Website URL
     3. Founder Name & Email
     4. Current Stage (e.g., Pre-Seed, Series A)
     5. One-sentence "Elevator Pitch"
   - *Once gathered, trigger the 'Get Lead' or 'Webhook' action.*

**KEY KNOWLEDGE:**
- **Firm:** [INSERT VC FIRM NAME].
- **Focus:** [INSERT FOCUS, e.g., Early Stage Fintech].
- **Check Size:** We write checks between $[INSERT MIN] and $[INSERT MAX].

**RULE:**
- Do not promise funding.
- Do not sign NDAs in the chat. (Refer to policy in `source`).

Step 3: The Knowledge Base (Tab: Source)

This is your “Investment Thesis.” The AI uses this to filter out the noise. If you only invest in Fintech, the AI will politely reject a Crypto pitch, saving you the email. Create *.txt file and copy paste text below:

### INVESTMENT THESIS
**Sectors We Love:**
- [INSERT SECTOR 1, e.g., B2B SaaS]
- [INSERT SECTOR 2, e.g., Artificial Intelligence]
- [INSERT SECTOR 3, e.g., Climate Tech]

**Sectors We AVOID:**
- Real Estate
- Brick & Mortar Retail
- Gambling / Adult

**Stage:**
- [INSERT STAGE, e.g., Pre-Seed and Seed].
- We like to be the first institutional check.

### CHECK SIZE & GEOGRAPHY
**Check Size:**
- Sweet spot: $[INSERT AMOUNT]k - $[INSERT AMOUNT]m.
- We do follow-on investments for top performers.

**Geography:**
- Primary: [INSERT REGION, e.g., North America].
- Secondary: [INSERT REGION, e.g., Europe].

### THE PROCESS
1. **First Screen:** Analysis of the deck (what this bot does).
2. **Founder Call:** 30-min Zoom with an Associate.
3. **Partner Meeting:** Deep dive with the decision makers.
4. **IC (Investment Committee):** Final decision.
5. **Term Sheet:** We issue the offer.

### FAQ
**Q: Do you sign NDAs?**
A: No, like most VC firms, we do not sign NDAs at the initial stage due to the volume of deals we see.

**Q: Do you lead rounds?**
A: [INSERT YES/NO].

### CONTACT
**General Email:** [INSERT EMAIL]
**Twitter/X:** [INSERT HANDLE]

Step 4: Enable Actions (Tab: Action)

Go to the Action tab.

Get Lead: This will email you the “Qualified Pitch” summary (Company, Stage, Pitch) instantly.
Webhook Call: Connect this to Airtable or Notion to automatically populate your Deal Flow pipeline with the new startup’s data.

Conclusion

An AI venture analyst helps you see more deals while reading fewer bad decks. It provides a better experience for founders (instant answers) and a cleaner pipeline for your partners.

Ready to automate your deal flow? Build your analyst agent now.

View Demo