Kruncher Logo
Build on Kruncher Β· For developers & platforms

The structured private-market layer your AI stack can reason on.

Use Kruncher as a building block, the layer beneath your product, not a UI you adopt. Call one primitive and compose it into your own application.

πŸ”’The model is rented. The data is the moat.
POST /v1/companies/analyze
# analyze any company against a thesis
{
  "company": "novacrop.com",
  "thesis_id": "agritech_seriesA"
}
# 200 OK
{
  "deal_score": 89,
  "thesis_match": true,
  "data_points": 1000,
  "sources": ["crunchbase", "linkedin", "registry"],
  "every_field": "timestamped + sourced"
}
The primitives

Five ways to build on the data layer

Structured, sourced, timestamped output, accessible programmatically, no UI required.

🏒

Company Enrichment API

Analyze any company against a defined thesis and receive structured, sourced, timestamped output.

πŸ‘€

People Enrichment API

Enrich any founder, executive, or professional into a structured, sourced profile.

πŸ”Œ

MCP Server

Connect Claude, ChatGPT, or custom agents directly to your data layer.

πŸ€–

Co-Worker (Agents)

Build your own agents to run repeatable tasks.

πŸ—„οΈ

Knowledge Graph Access

Query resolved entities and their relationships, companies, people, and funds. Shipped and visible in-product.

One primitive

Analyze a company or document against a thesis, get structured sourced output, compose it into your app.

Configurable by design

What the API can do

One call in, a structured, sourced view out, shaped the way your product and your thesis need it.

πŸ“„

Process PDFs & Excel

Send pitch decks, financial statements, and spreadsheets, and Kruncher parses and structures them.

🌐

Enrich with external data

Layer 20-plus premium data sources on top of the documents you send, resolved into one record.

βš™οΈ

Configure the output

Shape the response schema so it returns exactly the fields your product or workflow needs.

✍️

Add a custom prompt

Steer any analysis with your own instructions, passed in per request.

πŸ§ͺ

Define the analysis lenses

Set how market sizing, the team, and the competitive landscape are assessed.

🎯

Fit on 500+ signals

Score a company’s fit against 500 configurable signals, and add your own.

What it powers

Applications built on the primitive

Portfolio valuation & mark review

Re-analyze holdings on a schedule to support fair-value marks with sourced, timestamped evidence.

Secondary fund assessment

Assess a fund’s underlying companies fast when pricing an LP secondary or GP-led deal.

KYC / KYB & compliance validation

Validate companies and the people behind them against sourced, auditable records.

CRM enrichment at scale

Keep every company and contact record current across your CRM without manual data entry.

Corporate & competitive intelligence

Track competitors and markets continuously to feed corporate strategy and M&A teams.

Deal origination tooling

Power your own sourcing product with structured, on-demand company and people profiles.

Fintech fund Β· $14.5B AUM
Runs entirely on the API

Tracks thousands of companies with zero manual work, never touching the platform UI.

AI-native research team
MCP layer for a Claude workflow

Uses Kruncher as the structured data layer feeding its own Claude-powered research agents.

Build on the layer beneath the market.

One endpoint, structured and sourced, ready for your product or your AI stack.