NYAD AI Capital is a San Francisco-based early-stage venture fund with $180M AUM, backing the infrastructure companies that LLM applications, agents, and foundational AI systems depend on.
NYAD AI Capital is an early-stage venture fund with $180M AUM, founded in San Francisco in 2022. We invest in LLM infrastructure, AI agents, and foundational AI systems — the technical substrate that everything else in the AI ecosystem runs on. Pre-Seed through Series A, with a concentrated portfolio of ten to fifteen companies per fund.
We started NYAD because we kept seeing the same pattern: the real value in every technology cycle accrues to the infrastructure layer, not the applications on top. When cloud computing took off, the companies that endured were the databases, the compute layers, the networking primitives — not the first wave of SaaS apps. LLM inference, orchestration, data curation, evaluation frameworks — these are the picks and shovels of this cycle. They're harder to build, they compound more slowly, and they're worth dramatically more over time.
The founding team came from inside the AI research and engineering community — practitioners, not observers. That shapes how we evaluate companies, how quickly we move, and what we actually help with after the check clears.
We run a concentrated portfolio by design. Ten to fifteen companies per fund. We want enough time to be genuinely useful to each one.
How we think about AI bets and what we're actually looking for.
We chase the substrate. Every killer app creates ten infrastructure problems. The teams solving those problems are our targets.
The best AI companies form right at the edge of research and production. We invest before the market has priced that gap correctly.
We lead Pre-Seed and Seed rounds. We make decisions fast — usually inside two weeks — because that's what founders need at this stage.
We've made first-check decisions in 72 hours. Founders don't have time to wait while we run committee processes. We don't have committee processes.
We don't hedge into every deal. When we say yes, it's because we've done the technical work and we genuinely believe this is the right team for the problem.
We read the papers. We run the benchmarks. We talk to the people who would actually use what you're building. That's the diligence.
If you're building infrastructure that other AI companies will depend on, we want to hear from you. Send us a brief description of what you're building and we'll respond within 48 hours.
Get in Touch