Applied ML Scientist: RL for Data Center Optimization

Octaipipe; City Of London; Sur place; GBP 70 000 - 110 000; Plein temps

Résumé du poste

OctaiPipe, a dynamic UK-based AI hardware company, is seeking an Applied Scientist, Machine Learning to join the ACE team and work on reinforcement learning for data centre control. You will design, train, and deploy RL agents across live environments, translating telemetry into ML problems and collaborating with software engineers to productionise models on the OctaiPipe platform.

Candidates should have strong RL background, Python and ML tooling, and the ability to bridge research and

Qualifications

• An engineering (mechanical, electrical, control, systems) or physics degree at least level.
• Strong working knowledge of reinforcement learning with practical RL training/debugging experience.
• Solid Python and ML tooling (PyTorch/JAX, NumPy, RL libraries).
• Comfort with time-series sensor data and data quality issues.
• Ability to formulate operational problems as tractable ML problems and communicate trade-offs.
• Ability to iterate between research exploration and engineering deployment.

Responsabilités

• Design, train, and evaluate reinforcement learning agents for data centre control problems.
• Translate telemetry into well-posed ML problems, including state/action design and reward engineering.
• Sanity-check model behavior against physical first principles.
• Collaborate with software engineers to productionise models on the OctaiPipe platform (federated training, on-site inference, monitoring).
• Support the research agenda with academic collaborations and external write-ups.

Connaissances

Reinforcement learning

Python

PyTorch

JAX

Time-series data

Communication

ML tooling

Formation

Degree in engineering/physics

Outils

PyTorch

JAX

SciPy

Git

Description du poste

OctaiPipe, a dynamic UK-based AI hardware company, is seeking an Applied Scientist, Machine Learning to join the ACE team and work on reinforcement learning for data centre control. You will design, train, and deploy RL agents across live environments, translating telemetry into ML problems and collaborating with software engineers to productionise models on the OctaiPipe platform.

Candidates should have strong RL background, Python and ML tooling, and the ability to bridge research and