Build the immune system for agentic AI.
Autonomous AI systems are shipping into production faster than the security infrastructure to govern them. We are building that infrastructure. If that problem keeps you up at night, we should talk.
No patch Tuesday for autonomous agents.
We are architecting the control plane that will govern millions of autonomous systems: runtime behavioral analysis, threat detection, guardrail enforcement, governance at scale. The security gap is real, it is growing, and the window to close it is now.
Adversarial by default
To defend agents, you have to know how to break them. We think in attack surfaces, failure modes, and trust boundaries. We build systems that assume compromise, not systems that assume everything works.
No borrowed patterns
The security models built for request-response architectures do not map to autonomous, multi-step, tool-using agents. We are not adapting old frameworks. We are designing new primitives, informed by what came before but not constrained by it.
Roles we are hiring for.
Software Engineer
Architect the high-scale distributed systems that intercept, evaluate, and govern agent capabilities in real time. You will build the engine that makes the Trust Control Plane possible.
You are
- A distributed systems expert
- Obsessed with latency and throughput
- A polyglot (Go, Rust, TS)
Security Engineer
Define the threat models for the agentic era. You will build the simulation engines and runtime defenses that protect agents from prompt injection, tool abuse, and logic subversion.
You are
- A breaker and a builder
- Deeply knowledgeable in LLM security
- Comfortable with ambiguity
Agentic Security Researcher
Probe the boundaries of autonomous AI systems. You will design and execute adversarial experiments against agentic workflows, uncover novel failure modes, and turn findings into defensive primitives that ship into the platform.
You are
- Deep in LLM internals and agent architectures
- Published or active in agentic safety/security research
- Driven to turn research into production defenses
Applied ML Engineer
Build the ML models that power runtime behavioral analysis, anomaly detection, and trust scoring for autonomous agents. You will turn research into production-grade systems that operate at scale with strict latency requirements.
You are
- Experienced in production ML pipelines
- Strong in anomaly detection and behavioral modeling
- Comfortable shipping models with real-time constraints