AI
AI startups we scout
We look for AI companies that let organisations use their own data without surrendering it, private deployments and agents that finish real work.
Sovereign and agentic AI, before it becomes a category
Most AI pitches are a wrapper around someone else's model. We spend our time with the teams that own something harder: private deployment, data residency, agent reliability, auditability. We test their claims against real workloads, then introduce the ones that hold up to the enterprises and regulated operators asking us for exactly this.
- Proprietary data never leaves an environment the customer controls
- Agents that complete tasks with auditable steps and human checkpoints
- Real evaluation discipline instead of demo-day accuracy
Categories we track
Autonomous BI and knowledge systems
Teams making institutional knowledge queryable, with answers grounded in governed internal sources.
Private and sovereign GenAI
Startups running capable models inside customer-controlled environments, with residency and retention on the customer's terms.
Agentic workflow engines
Companies wiring task-completing agents into real business processes rather than chat windows.
Building in this space? Send us the product we test before we talk, and we'll tell you honestly where you stand.
Other focus areas
Security startups we scout
We track early-stage security companies that remove friction instead of adding it, agentless, identity-first, runtime-aware.
Read the thesisInfrastructure startups we scout
We back infrastructure teams giving operators a credible alternative to the incumbents, virtualization, containers and storage with predictable economics.
Read the thesisPartnerships
Partner with us
Founders: show us the product. Enterprises and operators: tell us the problem. Either way, the next conversation is a short one.