Ndea

Ndea has no commercial product and no revenue: it is an equity-funded research lab pursuing AGI through deep learning-guided program synthesis. The $43M raised — alongside participation in Y Combinator's W2026 batch — pays for a small, fully remote research team rather than hyperscale compute, which is consistent with the thesis itself: program synthesis bets that generalization comes from better abstractions, not bigger clusters, so the burn profile is structurally lighter than at scaling labs. The plan, to the extent one is public, is sequenced: first demonstrate systems that generalize from very few examples and adapt to novel tasks, then productize that capability. The founders' backgrounds hint at the eventual shape — Mike Knoop's Zapier history points toward automation that handles tasks no current agent generalizes to — but no pricing, customer, or go-to-market has been announced, and any revenue framing today would be speculation. For now, Ndea converts patient venture capital into research progress, full stop.