I reverse-engineered the tech stacks, valuations, and business models of the 30 private companies investors are most desperate to buy — Anthropic ($965B), OpenAI ($852B), Stripe, Anduril, Databricks + 25 more.
No paywall. No pitch deck. Each profile is a definitive guide: TL;DR, key facts, full tech stack, funding timeline, investors, pricing, traction, moat/risks, and 10+ hand-picked sources to go deeper — 300+ links total.
🔍 Same DB, searchable: startupwiki.tech · ✉️ I turn one of these into a free deep-dive every week: startupwiki.substack.com
⭐ If this saves you an hour of research, star it — it keeps the quarterly updates coming.
- AI eats the list: 19/30 are AI models, infra, or AI apps. The secondary market is basically an AI index fund now.
- $2T+ in private value: Top 3 alone (Anthropic $965B + OpenAI $852B + ByteDance $330B) outweigh the entire S&P bottom 300 combined.
- AWS wins, but nobody trusts one cloud: 18/30 primary on AWS, 6 on GCP, 4 on Azure — and every frontier lab is now multi-cloud + own data centers (Stargate, Rainier, Abilene).
- Python + TypeScript + Rust is the pre-IPO standard: Python for ML, TypeScript for product, Rust/Go for perf-critical infra. Ruby survives only at Stripe. C++ still rules robotics/defense.
- Prediction markets came out of nowhere: Polymarket ($15B) + Kalshi ($22B) went from fringe crypto to top-10 most-wanted on election + sports volume.
- Defense is the new SaaS: Anduril ($60B), Shield AI ($12.7B), Saronic ($9.3B), Hadrian — all with real revenue + government contracts, not just pilots.
- Vibe-coding is repricing dev tools: Cursor ($60B), Lovable ($6.9B), Replit ($9B) all crossed $100M ARR in <12 months.
Source: SetterVC Setter30 — quarterly ranking by buyer demand. Valuations = latest secondary marks.
| # | Company | Secondary | Cloud | Languages | Deep dive |
|---|---|---|---|---|---|
| 01 | Anthropic | $965B | AWS + GCP | Python, Rust, TS | Guide |
| 02 | Anduril | $60B | AWS GovCloud | C++, Rust, Python, Go | Guide |
| 03 | OpenAI | $852B | Azure + OCI | Python, TS, Go, Rust | Guide |
| 04 | Databricks | $134B | AWS + Azure + GCP | Python, SQL, Scala | Guide |
| 05 | Stripe | $159B | AWS | Ruby, Go, Java | Guide |
| 06 | Polymarket | $15B | AWS + Cloudflare | TS, Python, Solidity | Guide |
| 07 | ElevenLabs | $11B | GCP | Python, TS | Guide |
| 08 | Kalshi | $22B | AWS + Azure + GCP | Python, Go, TS, Rust | Guide |
| 09 | Neuralink | $9.7B | AWS + on-prem | Rust, Python, C++ | Guide |
| 10 | Crusoe | $10B | Own cloud + GCP | Python, Go, TS | Guide |
| 11 | ByteDance | $330B | OCI + AWS + own DCs | Go, Python, Java, C++ | Guide |
| 12 | Shield AI | $12.7B | AWS GovCloud + Azure Gov | C++, Python, Go | Guide |
| 13 | Revolut | $75B | GCP → AWS | Java, Kotlin, Python | Guide |
| 14 | Saronic | $9.3B | AWS GovCloud | C++, Rust, Python | Guide |
| 15 | Anysphere (Cursor) | $60B | AWS + Azure | TS, Rust, Python | Guide |
| 16 | Perplexity | $20B | AWS | Python, Go, Rust, TS | Guide |
| 17 | Ramp | $44B | AWS | Python, TS, Elixir | Guide |
| 18 | Project Prometheus | $41B | Undisclosed (stealth) | Python, C++, CUDA | Guide |
| 19 | Figure AI | $39B | Azure | Python, C++ | Guide |
| 20 | Hadrian | $1.6B | AWS (EKS) | Go, Python, TS | Guide |
| 21 | Lovable | $6.9B | AWS + GCP | TS, Go, Rust | Guide |
| 22 | OpenEvidence | $12B | GCP + Vercel | Python, TS | Guide |
| 23 | Harvey | $11B | Azure + AWS + GCP | Python, TS | Guide |
| 24 | Apptronik | $5.5B | AWS + GCP + NVIDIA edge | C++, Python | Guide |
| 25 | Physical Intelligence | $11B | GCP + AWS | Python, C++, Rust | Guide |
| 26 | Lightmatter | $4.4B | AWS | Python, C++, MATLAB | Guide |
| 27 | Canva | $42B | AWS | Java, TS, Python | Guide |
| 28 | Together AI | $7.5B | Own NVIDIA cloud | Python, TS, Go | Guide |
| 29 | Replit | $9B | GCP | JS/TS, Python, Go | Guide |
| 30 | Kraken | $20B | AWS + colo | Rust, Go, C++ | Guide |
Each guide = TL;DR + tech stack + funding timeline + investors + pricing + traction + moat/risks + 10+ sources.
├── companies/ # 30 definitive guides — TL;DR, tech stack, funding timeline, investors, pricing, traction, moat/risks, 10+ sources each
├── CONTRIBUTING.md # how to update for next quarter
└── LICENSE # MIT
Each company guide answers in 60 seconds, then goes deep:
- TL;DR + Key Facts — rank, valuation, founded, HQ, team size
- Full tech stack table — cloud, languages, DBs, infra, AI infra, frontend + architecture
- Funding timeline + cap table — rounds with dates, amounts, leads
- Business model + pricing — revenue streams with real $ tiers, GTM, margins
- Traction + moat/risks — revenue, users, growth, customers
- Learn More (10+ links) — official docs, must-read press, must-watch videos, engineering teardowns, data trackers
Example — Anthropic: AWS + GCP, Python/Rust/TS, Trainium2 + TPU serving, $965B secondary, $30B Series G at $380B, Google/Amazon/Spark/Menlo → API + subscriptions.
- Engineers: what stack to learn to get hired (spoiler: Python + TS + K8s + LLM APIs)
- Founders: how the best monetize (usage-based API + per-seat SaaS + marketplace)
- Job seekers / investors: which private companies have real traction vs hype
- Curious nerds: 300+ rabbit holes — Lex Fridman 8-hr Neuralink ep, Bloomberg's Anduril doc, Pragmatic Engineer's Cursor teardown
- Index = Setter30 Q2 2026 (buyer-demand ranking, published July 2026 — most recent as of Sept 2026).
- 6 parallel research agents cross-checked StackShare, BuiltWith, eng blogs, GitHub, job postings for stack; company announcements + Reuters/Bloomberg/TechCrunch/Forbes for rounds.
- Secondary valuations pinned to Setter marks; primary rounds dated. Stacks change fast — PRs welcome with sources.
- Not investment advice. Private marks are estimates and often differ from executable prices.
Found a mistake? Open a PR — corrections with a source get merged fast.
- Star it if you learned one new thing
- Share the table — every company links here
- PR next quarter's Setter30 when it drops (Q3 2026)
- Suggest the best talk/blog I missed in a company guide
I keep this updated quarterly and write up one startup deep-dive per week — business model, tech, and numbers, free.
- 🔍 Browse/search this DB: startupwiki.tech
- ✉️ Get the weekly deep-dive + Q3 update: startupwiki.substack.com
No spam, unsubscribe anytime. The repo will always stay free and MIT-licensed.
Built by StartupWiki — open startup research.
OSINT for learning only. Not investment advice. Valuations are secondary-market estimates, unaudited, and change quickly. Do your own diligence.
MIT © Contributors — see LICENSE.