Sources
Where the data comes from
Twenty primary sources, public and commercial. We document everything so you can verify, challenge, or replicate.
01 Primary data sources
| Source | What it provides |
|---|---|
| PitchBook | Venture deal flow + valuations (commercial) |
| Crunchbase | Startup funding, founder networks |
| CB Insights | Trend reports, market maps |
| USPTO Patent Search | Patent filings (public) |
| Google Patents | Global patent corpus |
| Bureau of Labor Statistics | Tech occupation employment |
| Census Business Builder | Local business density |
| LinkedIn Economic Graph | Talent flows, skill demand |
| GitHub Trending | Open source momentum |
| HuggingFace Models | AI model release velocity |
| arXiv cs.AI | Research publication rate |
| Austin Chamber | Local economic data |
| Greater Austin Inc. | Regional ecosystem reports |
| Austin Tech Council | Member directory + events |
| Tracxn | Sector taxonomies, emerging companies |
| Dealroom | Global startup database |
| OECD Science & Tech | Cross-country R&D indicators |
| World Bank Open Data | Macro tech indicators |
| Stanford AI Index | Annual AI ecosystem report |
| MIT Tech Review | Editorial trend coverage |
02 How we use them
- Commercial sources (PitchBook, Crunchbase, CB Insights, Tracxn, Dealroom) provide the bulk of capital data. We deduplicate across sources and prefer the highest-fidelity record.
- Government sources (BLS, Census, USPTO) provide ground truth on employment, business formation, and patents.
- Open data (arXiv, GitHub Archive, HuggingFace) provides the innovation signal.
- Local sources (Austin Chamber, Greater Austin Inc., Austin Tech Council) provide ground truth that big aggregators miss.
03 Data freshness
Capital data: weekly. Patent data: monthly. BLS employment: quarterly. The composite score is recalculated whenever any input updates.