Machine Learning Researcher

PRINCIPAL MACHINE LEARNING RESEARCHER (PHYSICAL AI)

Freeform builds AI-native manufacturing systems that unify software, hardware, and physics to produce industrial-scale parts at the speed of human ideation. By treating manufacturing as a single integrated system, we unlock a new era of innovation where complex hardware is designed, built, and scaled without limits.

This architecture enables continuous generation of petabyte-scale, high-fidelity data capturing the physics of metal printing - from in-situ process signals and machine state to geometry and material outcomes. Each factory node contributes to a growing learning system that improves modeling accuracy, control performance, yield, and scalability over time. 

Freeform is hiring a Principal Machine Learning Researcher to lead the development of advanced learning and control problems in a production-scale, AI-native metal manufacturing system. The role focuses on developing machine learning methods that integrate large-scale physical data with physics-based simulation and embedding these models into closed-loop control and autonomy frameworks. Work includes modeling relationships between process inputs, geometry, and machine state to predict thermal, mechanical, and geometric outcomes during printing, using hybrid physics–ML approaches and multi-modal in-situ data. 

Research is validated against physical outcomes and deployed into production systems, where improvements directly impact stability, yield, throughput, and capability across an expanding fleet of manufacturing nodes. Your work will have a direct and meaningful impact on how frontier technologies are designed and produced at scale. 

Responsibilities:

• Design and develop machine learning models for complex, multi-physics manufacturing processes. 
• Develop hybrid modeling approaches that combine first-principles physics with data-driven learning. 
• Lead the formulation of learning-based models used for prediction and control in production-scale metal additive manufacturing systems. 
• Develop methods to learn from large-scale, high-dimensional in-situ sensor data collected during printing. 
• Design unsupervised and self-supervised learning techniques to correlate process signals with part quality, geometry, and performance. 
• Develop models that link process parameters, geometry, and machine state to thermal and mechanical outcomes. 
• Integrate learned models with physics-based simulation and digital twin frameworks. 
• Contribute to the design of closed-loop control and autonomy systems that operate in real time on production hardware. 
• Develop learning-based approaches for machine health monitoring, anomaly detection, and system diagnostics.