/roles — ROLE_324
Member of Technical Staff [Platform]
Well-funded early lab building LLM agents that keep learning toward expert-level performance
The role
- COMP
- $200K - $300K
- EQUITY
- Competitive
- LOCATION
- South Bay Area
- WORKPLACE
- On-site
- EXPERIENCE
- 2 - 7 years
- VISA
- None, Visa transfers
- STACK
- AWS, GCP, Docker, Kubernetes, K8s, Terraform, CI/CD, Prometheus, Grafana
- INDUSTRY
- AI, Software Development
The company
Research lab developing specialized LLM agents that continually learn toward expert-level performance, founded by researchers whose agent work is widely used by frontier labs.
- STAGE
- growth-stage
- FUNDING
- $40M+ raised
- TEAM
- ~10 people
- FOUNDED
- 2025
JD — the work
About the role
You would create the in-house systems the lab runs on, powering both research experiments and shipped products. The role lives at the meeting point of MLOps, infra engineering, and dev tooling, and covers the serving layer for the agent: keeping the hosted agent available, secure, observable, and able to scale. You would work closely with researchers and engineers so experiments and releases are reproducible and efficient.
What you'll do
- Own the cloud compute, storage, observability, security, and CI/CD that research and product both depend on
- Write automation, shared libraries, and in-house tooling that make every team more productive
- Stand up dependable tooling to manage data and run, evaluate, and deploy models
- Shape how code gets built, tested, and released, so prototypes move smoothly into production
- Instrument services and pipelines with metrics, monitoring, and performance tracking
- Work with researchers and product engineers to find bottlenecks and make results reproducible
- Help plan infrastructure architecture for the lab's next stage of growth
What they're looking for
- 2 to 7 years of infrastructure, DevOps, or platform engineering experience
- Hands-on work with AWS or GCP, Docker, Kubernetes, and Terraform
- Experience running CI/CD and observability stacks such as Prometheus and Grafana
- Familiarity with MLOps workflows for experiments, evaluation, and model deployment