Cogent Security jobs

Cogent Security, an applied AI lab backed by Greylock, is hiring AI Fellows—early-career research engineers who've recently completed or are defending PhDs—to work on agentic security and auto-remediation within their Cogent Research team. You'll define and pursue your own research direction combining AI with formal methods (such as auto-formalization, certificate-generating code, runtime verification, or policy synthesis), moving ideas from prototype to production systems deployed with Fortune 500 customers. Required experience includes a strong publication record or comparable contributions in formal methods, program analysis, programming languages, agentic AI for code, or runtime verification, plus strong engineering ability and independent problem-solving skills in ambiguous settings. This is an early-career fellowship role with a California base salary of $100k–$300k plus potential equity and benefits, grounded in real-world security workflows rather than purely academic research. The standout detail is the explicit focus on combining AI agents with formal methods and non-bypassable guarantees—this isn't traditional security research but rather bridging two distinct technical areas at production scale.