Research Engineer, ML Infra for Scalable Training

Harvey; Greater London; On-site; GBP 100,000 - 160,000; Full time

Job summary

Harvey seeks a Member of Technical Staff, Infrastructure to help scale post-training stacks and RL rollout at the frontier of legal tech AI. You will own distributed training environments, tooling, and data management, collaborating closely with researchers to turn experiments into dependable production systems.

You’ll optimize GPU utilization, observability, and reliability to deliver fast, scalable models for real-world use, joining a small, high-bar team that values ownership and impact.

Qualifications

• Experience building ML infrastructure, distributed systems, or research tooling.
• Strong Python and backend software engineering skills with ML framework familiarity (PyTorch, JAX).
• Experience with distributed training or inference and diagnosing performance bottlenecks.

Responsibilities

• Scale our post-training stack, supporting SFT and RL on leading open-weight models.
• Build distributed rollout infrastructure coordinating model inference, tool execution, and training data collection.
• Create reproducible agent environments for training and evaluation with documents, tools, and external systems.
• Own performance and reliability of training runs, including GPU utilization, checkpointing, and observability.
• Identify and remove bottlenecks across training and inference to shorten research-to-result cycles.
• Develop internal tooling for data management, evals, results browsing, and tracking of training jobs and experiments.
• Integrate models into production, addressing serving latency, cost, and online evaluation.
• Collaborate with researchers to make experimental workflows dependable while allowing rapid experimentation.

Skills

ML infrastructure

Distributed systems

Python

Backend services

PyTorch

Tools

JAX

Job description

Harvey seeks a Member of Technical Staff, Infrastructure to help scale post-training stacks and RL rollout at the frontier of legal tech AI. You will own distributed training environments, tooling, and data management, collaborating closely with researchers to turn experiments into dependable production systems.

You’ll optimize GPU utilization, observability, and reliability to deliver fast, scalable models for real-world use, joining a small, high-bar team that values ownership and impact.