GitHub Usage Tips

GitHub signals surface engineering investment patterns from public repository activity. Rather than static technographic data ("uses Python"), these signals show active development direction — what companies are building, which technologies they're adopting, and how fast those projects are growing.

The data is delivered as structured JSON in weekly flat files.

Common Use Cases

• Technographic enrichment — Add real-time tech stack data to a B2B database, showing not just what companies use but where they're actively investing
• Product intelligence — Understand what products a company is building based on their public repositories and README content
• Buyer persona enrichment — Infer technical buyer personas from the technologies and frameworks a company is adopting
• Integration/partnership mapping — Identify which platforms and tools a company integrates with based on their SDK and plugin development

The example below walks through a common activation pattern with screenshots.

Example: Vector Database Prospecting

This example shows how a vector database company (Pinecone, Weaviate, etc.) would use GitHub signals to find qualified prospects.

The Filter Setup

• Signal Type: AI/ML Investment
• Confidence: High
• Technologies: RAG

This returns companies actively building RAG (retrieval-augmented generation) pipelines — Notion, Linear, Intercom, Shopify, Retool, Zapier — with live GitHub repos proving real investment.

Sample Signal

{
  "signal_id": "b8c3d912-5f6e-4c7b-8d9e-2a3b4c5d6e7f",
  "signal_type": "github-initiative",
  "signal_subtype": "githubAIMLInvestment",
  "association": "company",
  "company": {
    "name": "Intercom",
    "domain": "intercom.com",
    "linkedin_url": "https://linkedin.com/company/intercom",
    "industries": ["Software", "Customer Support", "SaaS"]
  },
  "data": {
    "summary": "Building Fin AI agent with massive knowledge base RAG",
    "detail": "Intercom's Fin repositories show heavy investment in retrieval-augmented generation for their AI support agent. LangChain integration with custom embeddings pipeline.",
    "relevance": 0.87,
    "confidence": "high",
    "sentiment": "positive",
    "technologies_mentioned": ["Python", "LangChain", "RAG", "Embeddings", "OpenAI"],
    "referenced_repos": ["fin-ai-agent", "knowledge-embeddings"],
    "portfolio_metrics": {
      "repository_count": 24,
      "growth": {
        "stars_pct": { "30d": 0.71, "60d": 0.89, "180d": 1.45 },
        "forks_pct": { "30d": 0.52, "60d": 0.68, "180d": 1.12 }
      }
    },
    "top_repositories": [
      {
        "name": "fin-ai-agent",
        "full_name": "intercom/fin-ai-agent",
        "url": "https://github.com/intercom/fin-ai-agent",
        "description": "RAG-powered AI agent for customer support",
        "current": { "stars": 890, "forks": 124 },
        "growth_pct": { "stars": { "30d": 0.71 }, "forks": { "30d": 0.52 } }
      }
    ]
  },
  "detected_at": "2026-01-24T10:15:00Z",
  "batch_id": "gh-20260124-def456"
}

What Makes This Useful

The technologies_mentioned field provides stack-level specificity. RAG, LangChain, Embeddings indicates the company needs vector storage infrastructure — not just that they're "doing AI."

The growth_pct_30d field (71% in this case) shows the project has momentum, distinguishing active investment from abandoned experiments.

Key Fields

Filter Dimensions

Broader Coverage Patterns

The example above uses narrow filters (AI/ML Investment + RAG technology) which surface highly qualified but lower-volume signals. For broader coverage, consider these patterns:

By Signal Type

By Field Combinations

Technology-based targeting — Use data.technologies_mentioned to filter by stack:

• Python, LangChain, OpenAI — AI/ML ecosystem
• Kubernetes, Terraform, Docker — Infrastructure
• TypeScript, React, Next.js — Frontend/fullstack

Growth-based targeting — Use portfolio_metrics.growth.stars_pct.30d to find momentum:

• > 0.5 (50%+ growth) — Rapid adoption, likely funded/prioritized
• > 0.2 (20%+ growth) — Active investment
• Any positive — At least not abandoned

Repository-based targeting — Use data.referenced_repos or top_repositories to find specific project types:

• SDK/plugin repos indicate platform plays
• Infrastructure repos (terraform-, k8s-) indicate ops investment
• AI repos (llm-, embeddings-, rag-*) indicate AI investment

Coverage vs. Precision Tradeoffs

Questions?

Contact [email protected] for integration support.