About OpenRouter

OpenRouter is the AI routing and infrastructure layer that AI builders, AI-native startups, and enterprises use to access, manage, and optimize their AI usage through a unified API, billing interface, and analytics platform. We route billions of tokens every month and sit at the center of how organizations operationalize LLMs across research, product, and production workloads.

We are a small team that punches above its weight. Every person here has direct impact on the product and our users.

About the Role

Most third-party risk roles hand you a mature program and ask you to keep the queue moving. This is the opposite.

You'll be the first security risk analyst at OpenRouter, building the vendor risk function from a blank page. The vendors you assess aren't the usual SaaS sprawl — they're the model providers and subprocessors sitting directly in our customers' data path. And you'll do it in a regulatory environment still being written: there's no playbook for how the EU AI Act applies to an AI routing layer and its supply chain. You'll help write ours.

If you've ever finished a vendor review and thought this should take a third as long and catch twice as much — and wanted to be the one to fix it — keep reading.

What You'll Do

• Own end-to-end security assessments for model providers, subprocessors, and SaaS tooling — and get vendors live without becoming the bottleneck.

• Read SOC 2 and ISO reports critically: scope, carve-outs, CUECs, exceptions, and whether the testing supports the opinion. Same for pen tests, DPAs, and subprocessor lists.

• Turn findings into decisions — residual risk and compensating controls, not a spreadsheet of yellow cells.

• Design and stand up the TPRM program: intake, tiering, SLAs, escalation, exceptions, and risk acceptance.

• Pitch and implement tooling that compresses time-to-close, integrated with our GRC stack (Drata) and ticketing.

• Build continuous monitoring for critical vendors and run annual reviews on a real cadence.

• Map vendor risk to our SOC 2, ISO 27001, HIPAA, GDPR, and EU AI Act obligations, including flow-down to subprocessors.

What We're Looking For

• 4+ years in third-party/vendor security risk or security assessment — real assessment reps, not just program administration.

• Working fluency across SOC 2, ISO 27001, HIPAA, and GDPR, plus enough command of the EU AI Act to reason about it rather than recite it.

• Technical literacy — cloud architecture, access models, encryption, data flows — enough to know when a vendor's answer doesn't hold up.

• Comfort with DPAs, BAAs, and security exhibits, and judgment about which clauses actually matter.

• A bias toward shipping. You'll pitch solutions and drive implementation yourself; nobody is going to manage your day.

• Clear writing and a high tolerance for ambiguity. When the precedent doesn't exist, you write the memo.

Nice to Have

• Experience assessing AI/ML vendors or inference infrastructure

• ISO 42001 or NIST AI RMF

• Scripting and automation to eliminate your own toil

• GRC platform administration (Drata, Vanta, or similar)

• Time at an early-stage startup where you built the function rather than joined it

• CISSP, CISA, CRISC, or CTPRP.

If you don't think you meet all of the criteria below but still are interested in the job, please apply. Nobody checks every box, and we're looking for someone who is excited to join the team.