← All work From the MZL portfolio · Security
Optimus
Security that runs continuously, driven by AI agents — so a security team of any size can operate like a much larger one.
The challenge
Security teams are outnumbered. Alerts pile up faster than people can read them, and the real signal hides in the noise. The job needed something that could watch, triage, and act around the clock — without adding headcount a team does not have.
What we built
An AI-native security layer where agents handle the continuous work: detecting, triaging, and responding in real time, escalating to people only when judgement is genuinely needed. The product is designed so the humans stay in command while the routine load lifts off them.
The role of AI
Intelligence is the core, not a feature. Agents are evaluated against real conditions, with clear boundaries on what they act on alone — the difference between a demo and something a team can trust in production.
Why it matters
It is the clearest expression of the Foundry approach: build the intelligence into how the product fundamentally works, then run it long enough to earn trust.
Built to be trusted in production — not demoed in a deck.