# OpenPitch
**Category:** π» Developer Tools
**Repository:** https://github.com/Avierovich/openpitch
**Views:** 0
**Installs:** 0
**Upvotes:** 0
**Directory Page:** https://allmcps.com/mcp/openpitch
## Description
Open, sourced, confidence-scored intelligence on AI startups - a free PitchBook alternative.
## Claude Desktop Quick Installation
Heuristic fallback β verify the package name and runner against the repository README before running it. Uses `npx` (confidence: low):
```json
"mcpServers": {
"openpitch": {
"command": "npx",
"args": ["-y","openpitch"]
}
}
```
## Documentation & README
# πͺ§ OpenPitch
**The open, real-time intelligence layer for AI startups β that any agent can build on.**
*A free, open-source alternative to PitchBook & CB Insights, focused on the AI companies VCs actually care about.*
`MCP-native` Β· `zero-cost` Β· `fully-sourced` Β· `updated daily`
[](https://github.com/Avierovich/openpitch/actions/workflows/ci.yml)
[](https://pypi.org/project/openpitch/)
[](https://pypi.org/project/openpitch/)
[](LICENSE)
> **Status: v0.1.3 β functional.** The pipeline, reconciliation engine, MCP server, and
> dashboard all work end-to-end. Coverage and source breadth keep growing via the daily run.
[**Browse the live dashboard β**](https://avierovich.github.io/openpitch/)
---
## Why OpenPitch exists
PitchBook and CB Insights cost **$20k+/year** β and for fast-moving AI startups, their data is often **months stale**, because human verification is slow. For a company growing 3Γ a year, a figure verified six months ago can be off by multiples.
Meanwhile, the real numbers are **already public**: founders state ARR on podcasts weeks before any database, funding hits SEC filings, hiring velocity reveals growth. They're just scattered, unstructured, and contradictory β exactly the problem an AI agent is built to solve.
**OpenPitch's bet is latency, not coverage.** For the AI companies that matter, a *fresh, fully-sourced, confidence-scored* number beats a *verified-but-stale* one. We don't claim certainty β we show you the receipts.
## What you get
Ask your coding agent, get an answer with receipts:
```
> what's Sierra's valuation, with sources?
Sierra β AI agents for customer service (sierra.ai)
Valuation $15.4B [consensus Β· confidence 0.96] Β· as of 2026-05
β³ 10 public sources Β· Reuters Β· CNBC Β· The Information Β· qz.com
β³ $950M round closed May 2026 β led by Tiger Global and GV
```
*(A real answer from the committed data β check it against the [live dashboard](https://avierovich.github.io/openpitch/).)*
Every number carries **its source, a confidence score, and a tracked history** of how it changed.
## Features
- ποΈ **Mines podcasts** β founders leak metrics on podcasts before any database catches them. We transcribe and extract them.
- π§Ύ **Always sourced** β every figure links to its origin (podcast timestamp, filing, article). No black-box numbers.
- π **Confidence-scored** β built from source reliability, speaker authority, corroboration, and freshness (confidence *decays* as data ages).
- π **Reconciles conflicts** β when sources disagree, you get a consensus range + a contradiction flag, not a silent guess.
- π§ **Learns which sources to trust** β sources that prove right over time earn more weight.
- π **Version-tracked** β the git history *is* the audit log. See exactly how a company's reported ARR evolved.
- π‘ **Composable** β emits typed events other agents subscribe to (newsletters, press alerts, investor outbound).
- π€ **A2A-discoverable** β ships an A2A agent card so agent ecosystems can find and describe it.
- π§― **Grounding** β give your AI a sourced, confidence-scored fact base so it stops making up AI-company numbers.
- β‘ **60-second install** β no key, no signup; works in your agent in under a minute.
- πΈ **Genuinely free** β runs entirely on free tiers. No cost to run, no cost to use.
## Quickstart β use it in Claude Code / Codex
**No API key. No signup. No cost.** The data is already built and committed; the MCP server just reads it, and *your* agent does the reasoning.
**Fastest β zero install** (reads the committed data from the public repo, no clone):
```bash
uvx openpitch-mcp
```
**Or install the package:**
```bash
pip install openpitch # the MCP server (mcp is a core dependency)
openpitch-mcp # start the read-only server
```
**Or run from a clone** (for the pipeline / to rebuild data):
```bash
git clone https://github.com/Avierovich/openpitch && cd openpitch
python -m venv .venv && source .venv/bin/activate
pip install -e ".[pipeline]" # core + pipeline LLM deps
openpitch seed # build the data/ database from the committed seed (offline, no key)
```
Then point your agent at the local server:
```jsonc
// MCP config (Claude Code / Codex) β zero-install via uvx:
{
"mcpServers": {
"openpitch": { "command": "uvx", "args": ["openpitch-mcp"] }
}
}
// (or "command": "openpitch-mcp" if you pip-installed the package)
```
Ask your agent: *"What's Cognition's ARR, with sources and confidence?"* β it calls `get_metric`/`get_provenance` and answers from committed data (and will flag the public-source discrepancy).
### Or just browse the data
- π **Live dashboard** β [avierovich.github.io/openpitch](https://avierovich.github.io/openpitch/) (sourced company cards, refreshed daily) β or build locally: `openpitch build-dashboard`
- π **Raw data** β [`data/companies/`](data/) β plain JSON, diffable, yours to use
- π€ **A2A Agent Card** β generated at `dashboard/dist/.well-known/agent.json`
> **Data status:** live, refreshed daily by CI. Figures are **probabilistic, public-source intelligence** β every number carries its source, confidence score, and date, and open quality items are [tracked in public](https://avierovich.github.io/openpitch/quality.html). See the [methodology](docs/METHODOLOGY.md) and the [correction workflow](docs/CORRECTIONS.md).
## Docs
- **Trust model** β [methodology](docs/METHODOLOGY.md) Β· [data policy](docs/DATA-POLICY.md) Β· [corrections](docs/CORRECTIONS.md)
- **Interfaces** β [MCP spec](docs/MCP-SPEC.md) Β· [events spec](docs/EVENTS-SPEC.md)
- **Architecture** β [full design doc](docs/FRD.md) Β· more product docs in [`docs/`](docs/)
## How it works
```
Sources Daily pipeline (free GitHub Actions) Interfaces
ββββββββββ βββββββββββββββββββββββββββββββββββ ββββββββββ
Podcasts ββ 1. select top-50 (VC-attention score) ββ MCP server (local, BYO agent)
News ββββββ€ ββββΆ 2. collect Β· 3. transcribe Β· 4. extract ββββΆ ββ static dashboard
SEC EDGAR β€ 5. reconcile Β· 6. score sources ββ event feed (JSONL)
Web βββββββ 7. publish β git commit (the database) ββ "what moved today" digest
```
The git repo **is** the database. There's no server to run. See the [FRD](docs/FRD.md) for the full design.
## Build on it (composability)
OpenPitch emits typed, confidence-scored **events** when something material changes β so other agents can react:
| You're building⦠| Subscribe to | OpenPitch becomes⦠|
|---|---|---|
| A newsletter agent | all material events | your content pipeline's data source |
| A press/PR workflow | funding/valuation events, confidence β₯ 0.8 | your "time to call the company" trigger |
| Investor outbound | universe entries, growth thresholds | your targeting signal |
Events ship on MCP and a raw `events/feed.jsonl`. Schemas are versioned. See the [events spec](docs/EVENTS-SPEC.md).
## How we compare
OpenPitch is **complementary to the incumbents, not a rip-and-replace.** We win a narrow wedge; we lose on breadth and verification β and we're honest about both.
| | PitchBook / CB Insights | Crunchbase | Harmonic | MAGNiTT / Wamda | **OpenPitch** |
|---|:--:|:--:|:--:|:--:|:--:|
| Price | $20kβ100k/yr | Freemium | Custom | $/regional | **Free & open** |
| Freshness | Weeksβmonths | Variable | Days | Weeks | **Daily** |
| In your AI agent (MCP) | β | β | β | β | **β** |
| Every figure sourced + confidence-scored | β | β | β | β | **β** |
| Contradiction detection | β | β | β | β | **β** |
| Coverage breadth | **βββ** | **βββ** | **ββ** | β (MENA) | narrow (by design) |
| Verified, diligence-grade | **β** | β | β | β | β (probabilistic) |
**The honest pitch:** *the free, fresh, AI-native first look β every number sourced β before you pull the expensive verified report.* For an investment decision, you still need the incumbents. Full mapping, feature matrix & pricing: [docs/COMPETITIVE-ANALYSIS.md](docs/COMPETITIVE-ANALYSIS.md) Β· [spreadsheet](docs/competitive-matrix.xlsx).
## Coverage
**Global AI startups** β **140+ profiled** across 12 sectors (including Chinese AI labs and European names Western trackers miss), with a **top 50 dynamically ranked** by VC attention (valuation + funding activity β *not* ARR, to avoid circularity). The list moves as attention shifts; companies entering/leaving the top 50 is itself a tracked signal, and auto-discovery grows the universe daily.
**MENA AI/tech segment** β a dedicated regional set (an open, AI-native alternative to MAGNiTT/Wamda). Honest caveat: MENA disclosure is lighter than the US, so this segment launches with lower confidence/coverage, clearly labeled.
Seed universe: [`config/watchlist.yaml`](config/watchlist.yaml).
## Honest disclaimer
OpenPitch is **transparently probabilistic**. Many figures are estimates derived from public, self-reported, sometimes-contradictory sources. We surface confidence and provenance precisely so you can judge for yourself. **This is not investment advice, and figures are not guaranteed accurate.** Always verify before acting.
## Roadmap
- [x] Seed universe (global AI + MENA segment) + auto-discovery (news, funding digests, 21-sector backfill, China feed)
- [x] Core data model + reconciliation engine (confidence, consensus, contradiction) β *tested*
- [x] Source adapters: podcast, news, EDGAR, company-site β *tested*
- [x] Extraction stage: batched LLM claim extraction + model rotation β *tested; data QA still required*
- [x] MCP server β local read-only data tools
- [x] Daily GitHub Actions pipeline β wired for LLM, Groq transcription, and SEC user-agent secrets
- [x] Static dashboard + company pages β generated from committed data
- [x] Event feed β JSONL feed and digest generated from publishes
- [x] A2A agent discovery card β generated with dashboard
- [ ] MENA adapters (regional news, free-zone registries)
- [ ] Rich-source expansion (GitHub, hiring, app-ranks) β *post-PMF scaling*
- [ ] *v2:* implied-ARR model, intra-day funding fast-lane
## Contributing
Contributions welcome β especially **new source adapters** (one file each) and **watchlist curation**. See the [FRD](docs/FRD.md) for architecture.
## Who built this
OpenPitch is built and run by [Mohamed Abdulhadi](https://www.linkedin.com/in/mohamedibnomer),
a product manager β working with AI agents (Claude Code) that wrote much of the code and now
operate the daily pipeline and its public data corrections. That's not a footnote; it's the
product demonstrating itself: an agent-native database, built and maintained agent-natively,
with every commit and correction in the open. Questions, feedback, or collaboration β
[connect on LinkedIn](https://www.linkedin.com/in/mohamedibnomer) or open an issue.
## License
[MIT](LICENSE)
---
Built in the open. Free forever. If a number looks wrong, open an issue β provenance means you can check our work.