Your investors' talent team, dedicated to your search.
VC funds have in-house recruiters — but they're spread across hundreds of portfolio companies. We bring the same network access and cap-table judgment, focused entirely on your hire.
Founding AI Engineer
Three ways to hire. One that actually reaches them.
Contingency recruiters send volume. AI sourcing tools stop at the top of the funnel. Your investors' own talent team is stretched across a hundred portfolio companies. We're one dedicated team with portfolio access and cap-table judgment — pointed entirely at your search.
Portfolio-grade access.
We recruit from inside our investors' portfolios — not cold-calling from the outside. That means candidates who aren't on the market — and a warm path to the ones who are.
Investor perspective.
We read cap tables, model equity outcomes, and workshop offers. Recruiting with founder-level context — the difference between a "maybe" and a signed offer.
Judgment you can't automate.
An embedded search partner who knows the candidates, understands the cap table, and closes with founder-level context.
Calibrate the real hire
We diagnose what the role actually needs — not a job description, the signal.
Work the network
We call people we know — and people who trust the people we know. Relationships, not databases.
Warm intros & references
Warm intros over spray-and-pray. Reference-backed signal before you invest time.
Founder-vetted shortlists
Founders see 3–5 people per search — each vetted through conversations and references, not keyword filters.
Close with investor context
Equity, leveling, comp — we help you frame the offer that actually closes.
What founders say.
Real outcomes from real searches. No aggregated testimonials, no names we can't share.
How the best AI companies get built.
Mirendil, Explained
Two Anthropic researchers want AI to do the job of two hundred AI researchers. Their thirteen identified people have done most stages of model training before, at DeepMind, Anthropic, xAI and the companies around them, and six are missing from the company's LinkedIn People page. The job board lists twelve open roles, two of them the gaps.
Flapping Airplanes, Explained
A year-old lab raised $1.18 billion for twenty people and publishes nothing. Its pitch is data efficiency; its record is GPU kernels. Index says a hire straight from high school built the agent infrastructure, Cursor runs through the founders' story, and the investors sit in the hiring loop. All twenty profiles read.
Latent Labs, Explained
An AlphaFold2 co-author took one engineer from the AlphaFold team, staffed a wet lab sixteen months before the first model, and made three protein-design launches in nine months with a team that, on our count, never passed seventeen. Twenty-six of 27 profiles read.
Get our frontier hiring research
Talent teardowns, comp signals, and how the best AI startups actually build their teams. New reports as we publish them — no spam, unsubscribe anytime.