Senior Research Infrastructure Engineer
Builds and owns the infrastructure, tools, and pipelines behind HUD's AI-agent research — reliable environment execution, trajectory collection, and large-scale model evaluation. Day to day is Python backend/infra work on cloud with Docker and Linux, partnering closely with researchers.
- Compensation
- $175K–$260K / yr
- Location
- San Francisco
- Work setup
- Onsite
- Seniority
- Senior
- Employment
- Full time
- Experience
- Not specified
- Category
- Devops
- Posted
Skills
Must have
- Strong software engineering fundamentals and proficiency in Python, with a record of building maintainable systems used by others
- Ownership of backend services, data pipelines, developer tools, or infrastructure from design through deployment and operation
- Sound judgment about APIs, data models, asynchronous workloads, and building reliable systems
- Comfort with Docker, Linux, cloud infrastructure, and debugging failures across application code and the systems it runs on
- Ability to work closely with researchers and translate open-ended research needs into useful software
- Independent technical judgment to set priorities and make pragmatic architecture decisions
Nice to have
- Working knowledge of reinforcement learning, including environments, rollouts, reward signals, and evaluation workflows
- Experience building research infrastructure, ML pipelines, or tools that made a research team faster
- Familiarity with distributed job execution, containerized environments, or high-concurrency workloads
- Track record of improving experiment reproducibility, data lineage, or visibility into complex pipeline failures
Full description
About HUD
HUD's mission is to build reliable, fair and open infrastructure for AI data. We want data to be valuable for the people who create it and trustworthy for the labs that train on it. Our team is a quickly growing group of researchers, engineers and operators building the economy that shapes what AI will become. Backed by $16M from top VCs and YC (W25), our marketplace and platform are used by startups, Fortune 500 companies and frontier labs.
About the role
We’re looking for a Senior Research Infrastructure Engineer to accelerate research on training and evaluating frontier AI agents. You’ll build the software and infrastructure that enable HUD’s researchers to iterate on environments, collect trajectories, and evaluate model behavior.
You’ll work closely with research engineers to understand the questions they’re investigating, identify bottlenecks in experimentation, and turn research code into reliable, reusable systems. You’ll own the tools, pipelines, and execution infrastructure that help researchers explore ideas, reproduce results, and scale promising approaches.
Responsibilities
- Build infrastructure that supports research on agent behavior through reliable environment execution, trajectory collection, and evaluation at scale
- Develop reusable tools, APIs, and pipelines that let researchers launch experiments, manage datasets and configurations, and inspect trajectories and evaluation results
- Turn research prototypes and experimental workflows into tested, documented software that researchers can reuse and extend
- Design experiments to understand model behavior, agent failure modes, and data quality issues
- Make experimentation faster, more reproducible, and more cost-efficient by improving orchestration, observability, and recovery from failed runs
- Partner with research and engineering to translate evolving research needs into infrastructure priorities and architecture decisions, owning systems from implementation through ongoing operation
ExperienceYou may be a good fit if you have:
- Strong software engineering fundamentals and proficiency in Python, with a record of building maintainable systems used by others
- Ownership of backend services, data pipelines, developer tools, or infrastructure from design through deployment and operation
- Sound judgment about APIs, data models, asynchronous workloads, and the tradeoffs involved in building reliable systems
- Comfort with Docker, Linux, cloud infrastructure, and debugging failures across application code and the systems it runs on
- The ability to work closely with researchers, understand experimental workflows, and translate open-ended research needs into useful software