Skip to content
Thesis

Our thesis is our filter — we prepare and source the companies rebuilding critical digital infrastructure.

Every layer of the digital economy — identity, code, models, networks, and money itself — is now contested terrain, and most of the tooling meant to defend it was designed for a slower, smaller attack surface. AI has changed the physics on both sides: it is the largest new attack surface in a generation and the most capable defensive primitive since the firewall. At the same time, finance is going programmable — payments, settlement, and real-world assets are being rebuilt on new rails that have to be secure and compliant from the first block. This is the view that guides who we take into the process: we prepare and source technical companies building security, infrastructure, and financial rails for that reality — the same companies we would one day want to back.

Where we focus
01

Securing the AI stack

As AI moves into production, it creates attack surfaces that legacy security tools were never built to see: autonomous agents that take actions, models that can be poisoned or extracted, and data pipelines with no provenance. We back teams building the identity, guardrails, and runtime controls for software that reasons and acts on its own.

Agent identity, authorization, and runtime governanceLLM/model red-teaming, evaluation, and runtime defenseData-poisoning and supply-chain integrity for training and RAG pipelines
02

Autonomous defense

The economics of security have inverted — attackers now hold the same generative tooling as defenders, while SOC teams can't hire their way out of alert volume. We fund applied-AI teams that compress detection, triage, and response from hours to seconds, and that treat the analyst as an operator of many agents rather than a queue of tickets.

Agentic SOC and autonomous alert triageAI-native detection engineering and threat huntingContinuous, autonomous offensive testing and red teaming
03

Critical digital infrastructure

The systems the economy actually runs on — machine identity, the software supply chain, cryptography, and the network fabric — are brittle, opaque, and increasingly targeted by nation-state and criminal actors. We back the foundational layer: infrastructure that is verifiable, resilient, and secure by default rather than by audit.

Machine and non-human identity at scaleSoftware supply-chain provenance, signing, and SBOMPost-quantum and cryptographic infrastructure
04

Applied AI for high-consequence work

In domains where a wrong answer is expensive — defense, finance, healthcare, industrial and public infrastructure — AI has to be reliable, auditable, and deployable inside real constraints. We back founders using AI to do consequential work in these environments, where trust, latency, and compliance are features and not afterthoughts.

AI for regulated and compliance-bound workflowsAutonomy and decision support for defense and industrial systemsVerifiable, auditable AI for finance and critical operations
05

Fintech & tokenization infrastructure

Money is becoming programmable, and the rails beneath it — payments, settlement, and the tokenization of real-world assets — are being rebuilt in the open. We back the teams building the secure, compliant infrastructure for that shift: the custody, identity, and market plumbing that let value move as freely as data without recreating the last generation's risks.

Tokenization of real-world and financial assetsPayments, settlement, and stablecoin infrastructureCustody, compliance, and on-chain identity rails
What makes a company a fit

The companies we take into the process.

  • Deep technical credibility — founders who have built, operated, or broken the systems they now want to defend, and whom strong engineers instinctively defer to.
  • Earned insight — a specific, non-obvious view of the problem that comes from having lived it as a practitioner, researcher, or operator inside the domain.
  • Clarity over narrative — the ability to explain hard things plainly; we are wary of teams who need jargon or hype to make the story stand up.
  • A sharp wedge — an acute first problem that pulls the product out of users' hands, with a credible path to owning the broader system over time.
  • Bias toward shipping — evidence they can move from research to production and get real users, not just demos, in front of the work.
  • Integrity under pressure — in a domain where trust is the product, we work with founders who tell us the bad news first.
How we work

The shape of a Sentinel engagement.

Stage

We work with companies raising from pre-seed through Series A and early growth rounds — from exceptional pre-product technical teams to companies with real traction sharpening their process before a larger raise.

Geographies

Primarily the U.S., U.K./Europe, and Israel — the deepest pools of security, infrastructure, and fintech talent. Remote-first teams welcome; we work with the right companies wherever they are.

Round size

Typically we prepare companies for raises from roughly €250K up to Series A and growth rounds of several million — the range where investor-grade preparation makes the clearest difference.

Thesis alignment

We prioritize companies that fit this long-term view. Alignment isn't a hard requirement to work with us, but it is how we build the proprietary, thesis-driven deal flow behind the fund we intend to become.

Apply

Ready to raise on institutional terms?

Start with an Investor Ready Audit. We'll show you exactly how an investor sees your company today — and what to fix before you go to market.