Read the company in market context
Compare the company with filings, patents, regulation, local networks, and changes over time.
- 01Check legal context
Review incorporation, filings, and regulatory exposure.
- 02Map the network
See cluster, talent, university, and customer links.
- 03Measure change
Compare trajectories instead of isolated snapshots.
Put every company in context.
Angel OS checks the startup inside its US regulatory, capital, procurement, talent, and IP environment before drawing a conclusion.
Resolve the legal entity, incorporation state, subsidiaries, Form D filings, funding vehicles, syndicates, and historical rounds before comparing the company.
A health startup is never analysed against generic global healthcare trends. It is analysed against the reimbursement path that decides whether anyone can pay for it.
Twelve clusters, ranked by graph density.
Start where company, founder, university, capital, and talent connections are dense enough to verify. National coverage remains on; dense clusters sharpen the first read.
SF Bay Area
Compiler, runtime and model-layer talent density is unmatched nationally.
- Tracked
- 4,820
- Adjacencies
- 8
- Graph weight
- 1.00
IP intelligence is the differentiator.
Check patents, trademarks, repositories, licences, and research. Determine what the IP protects and who owns it—not merely whether the company claims a moat.
- Does actual IP exist?
- What does it protect?
- Who is named as inventor?
- Was it assigned to the company?
- What is its remaining life?
- How broad are the claims?
- How easily can competitors design around it?
- Does it correlate with the shipped product?
- Is assignment recorded, or merely asserted?
- Did founders create related IP before incorporation?
- Is there university ownership risk?
- Could a former employer have a claim?
Resolve the three flagged questions first. They cover ownership conflicts, prior-employer exposure, and missing assignments.
The path matters more than the snapshot.
Track hiring, funding, products, and customers to reveal inflection points.
Eight monthly observations per series, all from public sources. Endpoint marked.