The Thesis

When the Gates Foundation quietly increases its climate-infrastructure holdings by 23% in a single quarter, that's not noise — it's signal. When Chan Zuckerberg Initiative and Berkshire Hathaway independently converge on the same biotech theme, that's not coincidence — it's conviction.

fin-sig is a capital intelligence platform I built to track these moments. It analyzes SEC 13F filings from mission-driven funders and scores each investment theme on three dimensions: confidence, novelty, and cross-funder convergence.

fin-sig app icon

Try fin-sig Live

Explore the full dashboard, live signals, API docs, and embeddable widgets.

Try the Dashboard API Docs
What are 13F filings? Every institutional investor managing over $100M must publicly disclose their equity holdings quarterly via SEC Form 13F. This is free, public data — fin-sig just makes it readable and actionable.

What fin-sig Does

Signal Scoring Engine

Every investment theme gets a composite score based on three factors:

  • Confidence (40%) — Data quality and position size relative to portfolio
  • Novelty (30%) — How recently the position changed (new entries score highest)
  • Convergence (30%) — How many independent funders are moving on the same theme

No AI calls. No black box. The scoring engine is deterministic Python math you can audit line by line.

Cross-Funder Convergence

This is the killer feature. When two or more unrelated funders — say, Gates Foundation and CZI — independently increase positions in the same sector, fin-sig flags it as a convergence event. These moments historically precede major public market repricing by 18–36 months.

Interactive Dashboard

A dark-themed editorial dashboard shows:

  • Funder cards with portfolio timeline charts
  • Click-to-expand quarter-over-quarter diffs (position changes, new entries, exits)
  • Signal cards with score breakdowns
  • Theme filters and watchlist management
  • One-click CSV export for your own analysis

Quarter-Over-Quarter Diffs

For each tracked funder, fin-sig computes what changed since last quarter: new positions, increased stakes, decreased positions, and full exits. You see conviction changes as they happen, not months later in a newsletter summary.

CSV Export & Open API

All data is exportable. Download signals or holdings as CSV for your own models, spreadsheets, or research. The REST API is documented with Swagger and ReDoc and available at no cost.

Embeddable Widgets

fin-sig ships embeddable widgets you can drop into any webpage with a single script tag. No build step, no framework dependency. Shadow DOM isolation means zero style conflicts with your existing site.

Available widgets:

  • Signal Card — Score gauge with sub-score breakdown and funder list
  • Convergence Badge — Inline badge showing multi-funder alignment
  • Timeline Sparkline — SVG portfolio value chart
<!-- Drop this anywhere on your site -->
<script type="module"
  src="https://jenninexus.com/fin-sig/widgets/finsig-widgets.js">
</script>

<finsig-signal-card
  data-theme="climate-infrastructure"
  data-api="https://jenninexus.com/fin-sig">
</finsig-signal-card>

Tracked Funders

Funder Source Key Themes
Gates Foundation Trust SEC 13F Global health, climate, education
Berkshire Hathaway SEC 13F Value investing, insurance, energy
Chan Zuckerberg Initiative SEC 13F Education, biomedical, housing
Emerson Collective Public portfolio Education, immigration, climate, media
Open Philanthropy Grant database AI safety, biosecurity, global health, animal welfare
Ford Foundation IRS 990-PF Social justice, democratic governance, economic equity

fin-sig tracks funders across multiple data sources — SEC 13F filings, public portfolio disclosures, grant databases, and IRS 990-PF filings — to build a complete picture of where mission-driven capital is flowing.

The Tech Stack

  • Backend: Python + FastAPI — lightweight, async-ready, auto-generated API docs
  • Data: SEC EDGAR API (free, public, no key required) with optional Supabase for persistence
  • Scoring: Pure Python — deterministic math, no external AI calls required
  • Dashboard: Single embedded HTML file — no build step, no framework
  • Widgets: Preact + esbuild — Shadow DOM, self-contained ESM bundle
  • Edge: Cloudflare Workers for cached API delivery at zero server cost

Running cost for the core platform: $0. SEC data is free. Python math is free. Local storage is free.

Building with AI — Anthropic & Claude

fin-sig was architected alongside Anthropic's Claude and the emerging financial analysis capabilities in the Claude ecosystem. A few resources worth exploring:

  • Anthropic Research — The reasoning capabilities behind financial document analysis
  • Claude Tool Use — Extending scoring engines with Claude-powered analysis tools
  • Claude — The AI assistant that helped architect fin-sig's data pipeline, scoring models, and dashboard design
Built with Claude: fin-sig's entire codebase — from the SEC EDGAR parser to the dashboard UI to the scoring engine — was developed in collaboration with Claude Code, Anthropic's CLI tool for software engineering.

Live Signal Preview

This is an embedded fin-sig widget. Visit the full platform page for the complete dashboard, all signals, and API access.

Explore All Signals on fin-sig

What's Next

The free dashboard, API, and embeddable widgets are live now. On the roadmap:

  • Additional data source integrations (grant databases, 990-PF filings)
  • Alert subscriptions for convergence events coming soon
  • AI-enhanced research briefs (optional, powered by Claude) coming soon
  • Team workspaces for collaborative research coming soon

If you're a researcher, developer, newsletter writer, or just curious about where smart money flows before the crowd — check out my GitHub for more projects like this.