Research Engineer, Infrastructure

Harvey

7d • $230,000 - $340,000

About The Position

As a Member of Technical Staff, Infrastructure in Harvey Research, you will build systems that power the adaptation and improvement of leading open-weight models for legal work. Harvey has post-trained its first model built for legal work and is building a team to scale its post-training efforts. You will own infrastructure for distributed training, RL rollout systems, model-learning environments for tool use and complex legal workflows, and internal systems for managing data, running evaluations, and launching and tracking training jobs. You will help determine what to build or adopt and how systems fit together, working on a small team of hands-on engineers and researchers. The role is suited to research engineers experienced in demanding ML or distributed systems, comfortable debugging across the stack, and motivated to work closely with researchers to make training faster, more reliable, and easier to iterate on. Harvey is hiring for this role across multiple levels.

Requirements

• Experience building ML infrastructure, distributed systems, or research tooling, with ownership of performance, reliability, and operation.
• Strong Python and generalist software engineering skills across backend services, data systems, and internal tools.
• Familiarity with PyTorch, JAX, or similar ML frameworks.
• Experience with distributed training or inference, including diagnosing performance bottlenecks and failures across the stack.
• Understanding of post-training workflows.
• Practical judgment about what to build versus adopt.
• Ability to partner with researchers and ship quickly as needs evolve.

Nice To Haves

• Experience with RL rollouts, agent environments, or sandboxed tool execution.

Responsibilities

• Scale the post-training stack to support supervised fine-tuning (SFT) and reinforcement learning (RL) on leading open-weight models.
• Build distributed rollout infrastructure coordinating model inference, tool execution, and collection of training experience for long-horizon tasks.
• Build reproducible agent environments where models interact with documents, tools, and external systems during training and evaluation.
• Own training-run performance and reliability, including GPU utilization, checkpointing, failure recovery, and observability.
• Identify and remove training and inference bottlenecks to shorten the path from research idea to useful result.
• Build internal tools and services for training data management and versioning, running evaluations and browsing results, and launching and tracking training jobs and experiments.
• Integrate models into production, including serving, latency, cost, and online evaluation, so research reaches customers quickly and safely.
• Partner with researchers to turn experimental workflows into dependable systems while allowing rapid experimentation.

What This Job Offers

• Job Type: Full-time
• Education Level: No Education Listed