Lead AI and Process Modelling Engineer at Alpha Machines

Job Title

Lead AI and Process Modelling Engineer

Not Disclosed + Equity

Alpha Machines is a London-based industrial AI company founded by repeat unicorn builders to re-industrialise Europe through advanced power electronics. They are building AlphaOS, a platform that combines physics-informed modelling with modern ML to design and manufacture critical propulsion systems for aerospace, defence, and robotics applications.

As a founding team member, you will own the AI and physics modelling layer of AlphaOS, bridging the gap between thermodynamics and machine learning. You will design prediction, anomaly-detection, and planning surfaces for digital twins, ensuring factory reasoning is grounded in physical truth to optimize yield and throughput for global customers.

London, UK

• Join a high-caliber founding team led by repeat unicorn founders with a proven track record of scaling massive technology companies.
• Work at the cutting edge of the "physical AI" revolution, building the software intelligence layer for vertically integrated European manufacturing facilities.
• Enjoy significant impact and ownership as a founding-team member with equity, defining the technical bar for how LLMs reason about physical systems.

• Architect the AI and physics modelling layer of AlphaOS, encompassing process-level prediction, anomaly detection, and root-cause analysis.
• Encode complex domain physics—including kinetics, thermodynamics, and chemistry—into the digital twin to drive production yield and quality.
• Design and ship LLM-driven reasoning surfaces, such as agents and retrieval systems, that produce actionable outputs for factory operators.

• Has 4+ years of experience in applied ML or computational physics, with a background at places like Deep Mind, Palantir, or top research groups.
• Possesses expert-level Python skills and fluency in PyTorch or JAX, with a track record of shipping production-grade ML systems.
• Demonstrates deep expertise in physics-informed neural networks, industrial time-series forecasting, or grounding LLM reasoning in structured operational data.