Our client, a well funded stealth AI for physics startup, is looking for a Member of Technical Staff on infrastructure.With roots at several of the world's leading universities and research institutes, they are building AI systems that can discover new physics at scale, sitting at the intersection of frontier AI research and fundamental science. This role owns the full infrastructure stack: cloud foundations, CI/CD, release engineering and production operations across a multi cloud environment. The immediate need is pure DevOps, with ML and GPU training operations a forward looking expansion rather than a day one requirement. Compensation is $250K to $350K base plus seed stage equity.Role & Responsibilities:- Own the full infrastructure stack end to end, from cloud foundations through CI/CD to production deployments- Build and operate multi cloud infrastructure for the AI platform across the existing and in progress cloud environments- Design and run the release engineering pipeline from commit to production, including automated tests, security scans and progressive rollouts- Establish infrastructure as code discipline, choosing tooling and making every research workflow and customer facing product deployable through code- Operate production infrastructure powering the paid API, model training jobs, agentic research workflows and customer deployments- Define and meet SLOs, build observability and alerting, schedule GPU and CPU capacity and lead incident responseSkills & Experience:- 4 to 6 years operating cloud infrastructure in production at companies known for engineering rigour, with founder exceptions considered- Deep fluency with infrastructure as code such as Terraform or Pulumi, plus modern CI/CD systems- Strong production operations background including observability, alerting and incident response- Multi cloud experience, with GPU and CPU capacity scheduling a plus- Bias toward building self service tooling that gives leverage to the wider engineering team- US citizen or green card holder strongly preferred, with sponsorship difficult for this role. Onsite, relocation required