Startup Ecosystem Intelligence MCP

> View on ApifyForge | Use on Apify Store

---

Quick Start

Add to your MCP client (Claude Desktop, Cursor, Windsurf):

{
  "mcpServers": {
    "startup-ecosystem-intelligence-mcp": {
      "url": "https://ryanclinton--startup-ecosystem-intelligence-mcp.apify.actor/mcp"
    }
  }
}

---

Startup ecosystem intelligence for VC deal sourcing gives your AI assistant instant access to 8 public data sources — patents, GitHub activity, job postings, ArXiv research, tech stacks, corporate registries, and SaaS competitive data — all fused into a single structured deal memo. Built for venture capitalists, corporate development teams, and accelerator managers who need quantified, behavior-based signals rather than self-reported pitch deck data.

This MCP server runs as a persistent Apify Standby actor and exposes 8 tools over the Model Context Protocol. Every tool call triggers parallel data collection across multiple APIs, applies one or more scoring algorithms (Innovation Velocity, Hiring Signal Decoder, Competitive Moat Analyzer, Corporate Health), and returns structured JSON with scores, signals, investment thesis points, and red flags. No manual research. No Crunchbase subscriptions. No waiting for founders to reply.

What data can you extract?

| Data Point | Source | Example |
|---|---|---|
| 📋 Corporate registration, entity status, officers | OpenCorporates (140+ jurisdictions) | "Acme Corp — Active, Delaware, 2 entities" |
| 🔬 US patent filings, assignees, claims | USPTO Patent Search | "7 patents filed, 3 in AI/ML classification" |
| 🌍 European patent applications | EPO Patent Search | "4 EP applications, 2 granted" |
| ⭐ GitHub repos, stars, forks, commit velocity | GitHub Repo Search | "23 repos, 4,812 stars, updated 3 days ago" |
| 🛠 Website tech stack components | Website Tech Stack Detector | "React, GraphQL, Kubernetes, Snowflake — 12 components" |
| 💼 Job postings by role, seniority, function | Job Market Intelligence | "34 open roles — 62% engineering, 18% sales" |
| 📄 Pre-publication ArXiv research papers | ArXiv Preprint Search | "9 papers, 5 published last 6 months" |
| 🏆 SaaS competitors, features, positioning | SaaS Competitive Intelligence | "8 direct competitors identified" |
| 🎯 Innovation Velocity Score (0-100) | Composite model | "Score: 74 — FAST velocity level" |
| 📊 Composite deal rating | All 8 sources | "DILIGENCE — compositeScore: 61" |

Why use Startup Ecosystem Intelligence MCP?

Manual startup research is slow, inconsistent, and biased toward companies with strong networks. A junior analyst takes 8-12 hours to gather patent data, check corporate registries, scan job boards, and benchmark GitHub activity for a single company. Multiplied across a pipeline of 50 deals per month, that is 400-600 analyst hours just to screen at the top of the funnel.

This MCP automates the entire observable-signal layer of startup due diligence. Connect it to Claude, Cursor, or any MCP-compatible client and run a full deal memo in under two minutes.

- Standby mode — the server stays warm and responds instantly without cold-start delays between queries
- API access — call any of the 8 tools programmatically from Python, JavaScript, or curl
- Parallel data collection — up to 8 Apify actors fire simultaneously, so a deal memo takes 60-90 seconds rather than running sequentially
- Spending controls — set a per-session budget; the server returns a clean error when the limit is reached
- Integrations — connect to Zapier, Make, or webhooks to trigger deal memo generation from a CRM deal stage change

Features

- Innovation Velocity Score (0-100) — weighted composite of USPTO patents (up to 25 pts), EPO patents combined, GitHub repo count (up to 15 pts), total GitHub stars on a log2 scale (up to 15 pts), ArXiv paper count (up to 25 pts), and a recency bonus for activity in the past 6 months (up to 20 pts)
- Velocity level classification — five tiers: DORMANT, SLOW, MODERATE, FAST, HYPERGROWTH based on composite score thresholds (0/20/40/60/80)
- Hiring Signal Decoder — classifies 34+ job title keywords into 6 role categories (engineering, sales, marketing, executive, operations, other) and infers strategic direction: BUILDING (≥50% engineering), SCALING (≥40% sales), PIVOTING (≥30% executive), or MAINTAINING
- Competitive Moat Analyzer — scores tech stack depth (up to 25 pts), patent protection (up to 30 pts), competitor density using an inverse scoring model (fewer competitors = higher score), and GitHub star community proxy as network-effect signal (up to 20 pts via log2 scaling)
- Moat type classification — five levels: NONE, WEAK, MODERATE, STRONG, FORTRESS
- Corporate Health Check — active entity ratio scoring, existence verification, structural complexity penalty (entities beyond 3 incur penalty), and jurisdiction scoring that rewards 1-3 jurisdictions over tax-haven complexity
- Deal rating engine — composite score weighted as Innovation 30% + Moat 25% + Hiring 25% + Corporate 20%; maps to PASS / WATCH / DILIGENCE / STRONG_BUY at 25/50/75 thresholds
- Red flag detection — auto-surfaces concerning signals: dormant innovation, weak moat in a dense market, ≥3 dissolved entities, poor corporate health
- Investment thesis generation — auto-generates 1-4 thesis bullet points when positive signals cross thresholds
- 8 specialized MCP tools — each independently callable for targeted analysis without running the full deal memo
- Parallel actor orchestration — runActorsParallel() uses Promise.allSettled() so a single failing data source does not block the entire analysis
- Jurisdiction filtering — discover_startups and verify_corporate_structure accept an optional jurisdiction code for country-specific corporate registry queries

Use cases for startup ecosystem intelligence

VC deal sourcing with alternative data

Venture associates processing 40-80 inbound decks per week need a fast triage layer before partner time. Use discover_startups to surface active companies in a sector, then generate_deal_memo to score the top candidates on observable behavior — patent velocity, GitHub credibility, hiring direction — before scheduling calls. Replace the first week of due diligence with a 90-second automated screen.

Corporate development and acquisition scouting

Corporate development teams at technology companies need to map IP landscapes before approaching acquisition targets. analyze_competitive_moat reveals patent portfolio depth and tech stack complexity for any named company. assess_innovation_velocity surfaces which players are accelerating R&D output before they become expensive. Run this across a watchlist of 20 companies weekly with scheduled MCP calls.

Accelerator and portfolio benchmarking

Accelerator operators managing cohorts of 20-40 companies need a consistent evaluation framework. benchmark_against_cohort compares a portfolio company's tech stack maturity, open hiring positions, and competitive density against the sector. decode_hiring_signals monitors whether a portfolio company has shifted from building to scaling mode — an early signal of product-market fit.

Technology trend scouting and thesis development

LP-facing investment teams building thesis documents need to quantify technology momentum before capital allocation. track_technology_trends queries ArXiv paper counts, GitHub star velocity, and USPTO patent filings for a technology area (e.g., "diffusion models", "vector databases", "solid-state batteries") and returns a structured signal picture from research to commercialization.

Pre-investment due diligence automation

Legal and finance teams conducting formal due diligence can use verify_corporate_structure to check entity status, jurisdiction count, and inactive entity history across 140+ corporate registries via OpenCorporates before paying for a law firm to run the same check. Flags like dissolved subsidiaries or unusual jurisdiction stacking surface immediately.

Competitive landscape mapping for portfolio companies

Portfolio company founders preparing competitive analysis for board presentations can use benchmark_against_cohort to enumerate direct competitors, compare tech stack sophistication, and quantify open hiring volume as a proxy for competitor growth rate.

How to use startup ecosystem intelligence

1. Connect the MCP server — Add the endpoint https://startup-ecosystem-intelligence-mcp.apify.actor/mcp to your MCP client (Claude Desktop, Cursor, Windsurf, or Cline) with your Apify API token as the Bearer token.
2. Choose your tool — Ask your AI assistant to run a deal memo ("Generate a deal memo for Cohere") or a targeted analysis ("What is Databricks' hiring strategy right now?").
3. Wait 60-90 seconds — The server queries up to 8 data sources in parallel. Full deal memos take longer than single-dimension tools.
4. Review the structured output — Scores, ratings, signals, investment thesis points, and red flags are returned as structured JSON that your AI assistant can reason over and summarize.

Input parameters

This is an MCP server. There are no actor-level input parameters — connection is handled by your MCP client. Each tool accepts its own arguments as described below.

Tool parameters

| Tool | Parameter | Type | Required | Description |
|---|---|---|---|---|
| discover_startups | query | string | Yes | Technology, market sector, or keyword to search |
| discover_startups | jurisdiction | string | No | Country/jurisdiction code filter (e.g., "us_de", "gb") |
| assess_innovation_velocity | company | string | Yes | Company or organization name |
| decode_hiring_signals | company | string | Yes | Company name |
| decode_hiring_signals | location | string | No | Location filter for job postings |
| analyze_competitive_moat | company | string | Yes | Company name |
| analyze_competitive_moat | website | string | No | Company website URL — enables tech stack detection by URL rather than name |
| verify_corporate_structure | company | string | Yes | Company name |
| verify_corporate_structure | jurisdiction | string | No | Jurisdiction code to filter corporate registry results |
| track_technology_trends | technology | string | Yes | Technology, framework, or research topic |
| benchmark_against_cohort | company | string | Yes | Company name to benchmark |
| benchmark_against_cohort | website | string | No | Company website URL |
| generate_deal_memo | company | string | Yes | Startup company name |
| generate_deal_memo | website | string | No | Company website URL — improves tech stack detection accuracy |

Connection examples

Claude Desktop — add to claude_desktop_config.json:

{

  "mcpServers": {

    "startup-ecosystem": {

      "url": "https://startup-ecosystem-intelligence-mcp.apify.actor/mcp",

      "headers": {

        "Authorization": "Bearer YOUR_APIFY_TOKEN"

      }

    }

  }

}

Cursor or Windsurf — MCP settings:

{

  "startup-ecosystem-intelligence": {

    "url": "https://startup-ecosystem-intelligence-mcp.apify.actor/mcp",

    "headers": { "Authorization": "Bearer YOUR_APIFY_TOKEN" }

  }

}

Direct HTTP call:

curl -X POST "https://startup-ecosystem-intelligence-mcp.apify.actor/mcp" \

  -H "Content-Type: application/json" \

  -H "Authorization: Bearer YOUR_APIFY_TOKEN" \

  -d '{"jsonrpc":"2.0","method":"tools/call","params":{"name":"assess_innovation_velocity","arguments":{"company":"Mistral AI"}},"id":1}'

Input tips

- Provide the website URL when available — analyze_competitive_moat and generate_deal_memo use it for tech stack detection by URL, which is more accurate than name-based lookup.
- Use jurisdiction codes for corporate verification — Pass "us_de" for Delaware, "gb" for UK, "de" for Germany. Without it, results include all matching jurisdictions.
- Run assess_innovation_velocity first for pre-revenue companies — patents and ArXiv papers are available even when revenue data is not.
- Use track_technology_trends for thesis building — Query a technology area ("agentic RAG", "model distillation") rather than a company name to map the research-to-commercialization pipeline.

Output example

Full output from generate_deal_memo for a hypothetical company:

{
  "company": "NovaSynth AI",
  "compositeScore": 68,
  "dealRating": "DILIGENCE",
  "innovationVelocity": {
    "score": 74,
    "patentCount": 5,
    "epoPatentCount": 3,
    "githubRepos": 18,
    "githubStars": 6240,
    "arxivPapers": 7,
    "velocityLevel": "FAST",
    "signals": [
      "8 patents filed (5 USPTO, 3 EPO)",
      "6240 GitHub stars across 18 repos — strong developer community",
      "18 public repositories — active open source presence",
      "7 ArXiv publications — research-driven innovation"
    ]
  },
  "hiringSignals": {
    "score": 61,
    "totalJobs": 28,
    "engineeringJobs": 16,
    "salesJobs": 5,
    "executiveJobs": 2,
    "strategyInference": "BUILDING",
    "roleDistribution": {
      "engineering": 16,
      "sales": 5,
      "marketing": 3,
      "executive": 2,
      "operations": 1,
      "other": 1
    },
    "signals": [
      "57% engineering hires — product building phase",
      "28 open positions — significant growth"
    ]
  },
  "competitiveMoat": {
    "score": 58,
    "techStackDepth": 14,
    "competitorCount": 6,
    "patentProtection": 8,
    "moatType": "MODERATE",
    "moatFactors": [
      "Technical complexity",
      "Patent portfolio"
    ],
    "signals": [
      "14 technology components detected — complex tech stack",
      "8 patents providing IP protection"
    ]
  },
  "corporateHealth": {
    "score": 82,
    "entityCount": 2,
    "activeEntities": 2,
    "inactiveEntities": 0,
    "jurisdictions": ["us_de", "gb"],
    "healthLevel": "STRONG",
    "signals": [
      "2 corporate registration(s) found"
    ]
  },
  "allSignals": [
    "8 patents filed (5 USPTO, 3 EPO)",
    "6240 GitHub stars across 18 repos — strong developer community",
    "7 ArXiv publications — research-driven innovation",
    "57% engineering hires — product building phase",
    "28 open positions — significant growth",
    "14 technology components detected — complex tech stack",
    "2 corporate registration(s) found"
  ],
  "investmentThesis": [
    "High innovation velocity (74/100) — strong R&D output",
    "Engineering-heavy hiring — product building phase (ideal for early-stage)"
  ],
  "redFlags": []
}

Output fields

| Field | Type | Description |
|---|---|---|
| company | string | Company name as provided |
| compositeScore | number (0-100) | Weighted composite: Innovation 30% + Moat 25% + Hiring 25% + Corporate 20% |
| dealRating | string | PASS / WATCH / DILIGENCE / STRONG_BUY at thresholds 25/50/75 |
| innovationVelocity.score | number (0-100) | Innovation Velocity Score |
| innovationVelocity.patentCount | number | USPTO patents found |
| innovationVelocity.epoPatentCount | number | EPO patents found |
| innovationVelocity.githubRepos | number | GitHub repositories found |
| innovationVelocity.githubStars | number | Total stars across all repos |
| innovationVelocity.arxivPapers | number | ArXiv preprints found |
| innovationVelocity.velocityLevel | string | DORMANT / SLOW / MODERATE / FAST / HYPERGROWTH |
| innovationVelocity.signals | string[] | Human-readable evidence statements |
| hiringSignals.score | number (0-100) | Hiring Signal Score |
| hiringSignals.totalJobs | number | Total open positions found |
| hiringSignals.engineeringJobs | number | Engineering/developer roles |
| hiringSignals.salesJobs | number | Sales/BD/revenue roles |
| hiringSignals.executiveJobs | number | VP/Director/C-suite roles |
| hiringSignals.strategyInference | string | BUILDING / SCALING / PIVOTING / MAINTAINING / UNKNOWN |
| hiringSignals.roleDistribution | object | Count per role category (engineering, sales, marketing, executive, operations, other) |
| competitiveMoat.score | number (0-100) | Competitive Moat Score |
| competitiveMoat.techStackDepth | number | Technology components detected |
| competitiveMoat.competitorCount | number | Direct competitors found |
| competitiveMoat.patentProtection | number | Total patents (USPTO + EPO) |
| competitiveMoat.moatType | string | NONE / WEAK / MODERATE / STRONG / FORTRESS |
| competitiveMoat.moatFactors | string[] | Named moat sources (e.g., "Patent portfolio", "Community/network effects") |
| corporateHealth.score | number (0-100) | Corporate Health Score |
| corporateHealth.entityCount | number | Total corporate entities found |
| corporateHealth.activeEntities | number | Active/good standing entities |
| corporateHealth.inactiveEntities | number | Dissolved/revoked entities |
| corporateHealth.jurisdictions | string[] | Unique jurisdiction codes found |
| corporateHealth.healthLevel | string | POOR / CONCERNING / ACCEPTABLE / GOOD / STRONG |
| allSignals | string[] | All signal statements from all four models combined |
| investmentThesis | string[] | Auto-generated positive thesis points |
| redFlags | string[] | Auto-generated concern statements |

How much does it cost to run startup due diligence?

This MCP uses pay-per-event pricing — you pay $0.045 per tool call. Platform compute costs are included. The generate_deal_memo tool runs 8 actors in parallel but still costs a flat $0.045.

| Scenario | Tool calls | Cost per call | Total cost |
|---|---|---|---|
| Quick test — single innovation check | 1 | $0.045 | $0.045 |
| Targeted analysis — 3 dimensions | 3 | $0.045 | $0.14 |
| Full deal memo — one company | 1 | $0.045 | $0.045 |
| Screen 10 companies with deal memos | 10 | $0.045 | $0.45 |
| Full pipeline — 50 deals screened | 50 | $0.045 | $2.25 |

You can set a maximum spending limit per session in your Apify account settings to control costs. The server returns a clean error message when the budget is reached so your AI client can report it gracefully.

Compared to PitchBook at $20,000+/year or CB Insights at $6,000+/year, most VC teams using this MCP for deal screening spend under $20/month with no subscription commitment and no per-seat pricing.

Using startup ecosystem intelligence via the API

Python

import requests
import json
response = requests.post(

    "https://startup-ecosystem-intelligence-mcp.apify.actor/mcp",

    headers={

        "Content-Type": "application/json",

        "Authorization": "Bearer YOUR_APIFY_TOKEN",

    },

    json={

        "jsonrpc": "2.0",

        "method": "tools/call",

        "params": {

            "name": "generate_deal_memo",

            "arguments": {

                "company": "Cohere",

                "website": "https://cohere.com"

            }

        },

        "id": 1

    }

)
result = response.json()

memo = json.loads(result["result"]["content"][0]["text"])
print(f"Company: {memo['company']}")

print(f"Composite Score: {memo['compositeScore']}/100")

print(f"Deal Rating: {memo['dealRating']}")

print(f"Innovation Velocity: {memo['innovationVelocity']['score']}/100 ({memo['innovationVelocity']['velocityLevel']})")

print(f"Hiring Strategy: {memo['hiringSignals']['strategyInference']} — {memo['hiringSignals']['totalJobs']} open roles")

print(f"Moat: {memo['competitiveMoat']['moatType']}")

print(f"Red Flags: {memo['redFlags']}")

JavaScript

const response = await fetch(
  "https://startup-ecosystem-intelligence-mcp.apify.actor/mcp",
  {
    method: "POST",
    headers: {
      "Content-Type": "application/json",
      "Authorization": "Bearer YOUR_APIFY_TOKEN",
    },
    body: JSON.stringify({
      jsonrpc: "2.0",
      method: "tools/call",
      params: {
        name: "assess_innovation_velocity",
        arguments: { company: "Mistral AI" },
      },
      id: 1,
    }),
  }
);
const result = await response.json();

const velocity = JSON.parse(result.result.content[0].text);
console.log(Innovation Velocity Score: ${velocity.innovationVelocity.score}/100);

console.log(Level: ${velocity.innovationVelocity.velocityLevel});

console.log(Patents: ${velocity.innovationVelocity.patentCount} USPTO + ${velocity.innovationVelocity.epoPatentCount} EPO);

console.log(GitHub: ${velocity.innovationVelocity.githubRepos} repos, ${velocity.innovationVelocity.githubStars} stars);

console.log(ArXiv: ${velocity.innovationVelocity.arxivPapers} publications);

cURL

```bash