Investors for AI agent and applied AI startups
Most new software is now built around models. We back AI that does a defined job for a business buyer, priced on the work it does, and measured in customer outcomes. We also back the tools agents need in production: evaluation, security and orchestration.
What we look for
- An agent or product that completes one workflow end to end, not a demo
- Paid usage that continues after the pilot, and retention by cohort
- Gross margin after inference costs, and a plan for how it improves with scale
- A moat beyond the model: proprietary data, deep workflow integration or distribution
Usually not a fit
- Thin wrappers around a model with no data, workflow or distribution advantage
- Products a model provider can ship as a feature next quarter
Check size: $100k–500k through our syndicate. Stage: pre-seed and seed, after first revenue. Full criteria.
Frequently asked questions
Do you need SaaS-level gross margins?
No. AI products often start at 50–60% gross margin. We want to see the margin after inference costs and how it rises as you scale.
Do you invest in AI infrastructure and developer tools?
Yes, when developers use and pay for the product. Evaluation, security, memory and orchestration for agents are all in scope.
Do you invest before revenue?
Rarely. AI companies can reach first revenue fast, so we expect paying customers or signed paid pilots.
Raising for your AI agents and applied AI startup?
Send us your deck and a few numbers. It takes about five minutes.
Every submission gets a reply within two days.
Submit your deck