Applied Scientist, Machine Learning

We are looking for an Applied Scientist, Machine Learning to join the ACE team and work on reinforcement learning applied to real, physical infrastructure. You will help develop, train, and harden RL agents that operate in live data centre environments, working across the full arc from problem formulation and model training through to federated deployment and inference on customer sites. The work sits at the intersection of machine learning and engineering: you will spend as much time reasoning about thermodynamics, equipment behaviour, and operational constraints as you do about model architectures and training dynamics.

Duties and Responsibilities

• Design, train, and evaluate reinforcement learning agents for control problems in data centre environments.

• Translate messy, real-world telemetry into well-posed ML problems, including state and action design, reward engineering, and constraint handling for safety-critical operation.

• Sanity-check model behaviour against physical first principles and catch unrealistic results before they propagate downstream.

• Work alongside software engineers to productionise models on the OctaiPipe platform, including federated training pipelines, on-site inference, and monitoring of deployed policies.

• Support the team’s research agenda, including collaborations with academic partners and (where appropriate) external technical write-ups.

• An engineering (mechanical, electrical, structural, control, chemical, systems, or similar), physics or similar background at degree level or above.

• Strong working knowledge of reinforcement learning, including practical experience training and debugging deep RL agents on non-trivial problems

• Solid Python and modern ML tooling (PyTorch or JAX, NumPy, common RL libraries)

• Comfort working with time-series sensor data, including the realities of missingness, drift, calibration issues, and noisy labels.

• Ability to formulate ambiguous operational problems as tractable ML problems, and to communicate the resulting trade-offs to both technical and non-technical stakeholders.

• Comfortable iterating between research-style exploration and the engineering work needed to get something running on a real site.

You also might have

• Direct experience with control systems (classical control, MPC) or with HVAC, thermodynamics, power systems, or data centre operations.

• Experience with simulation, building or using physics-based simulators, digital twins, surrogate models, or large physics models.

• Familiarity with graph neural networks, meta-learning, multi-task learning, or offline/safe RL.

• Experience with federated learning, distributed training, or edge ML deployment.

• A track record of published research, open-source contributions, or relevant industrial RL deployments.

• Exposure to carbon-aware computing, demand response, or sustainability-driven optimisation.

• Postgraduate qualifications (MSc/PhD) in a relevant engineering, physics or ML discipline.

OctaiPipe is a young, ambitious company with the vision to be the global driving force for the next paradigm of foundational, physical AI that ensures our connected world, and its critical infrastructure, is safe, secure and sustainable. We are growing fast, having closed a recent funding round and looking to accelerate rapidly. OctaiPipe is offering the right candidate an exciting role on this adventure!

OctaiPipe is on a mission to revolutionise the optimisation of energy in data centres through decentralised artificial intelligence (AI). To do this, OctaiPipe is harnessing an elegant but revolutionary idea. Rather than move the data from the source, move the algorithms to the data to learn at the data source. This learning can be achieved with the intelligence of many devices through novel federated AI technology. OctaiPipe is developing the AI for Cooling Efficiency (ACE) application to be deployed using its own in-house distributed AI platform.