Research Engineer - Thermodynamic Computing jobs in United States
cer-icon
Apply on Employer Site
company-logo

Noumenal Labs · 1 month ago

Research Engineer - Thermodynamic Computing

Noumenal Labs is a deep tech AI company focused on closing performance gaps in outdoor robotics. As a Research Engineer, you will shape the technical architecture and research direction for thermodynamic computing, designing and deploying hybrid computing stacks to enhance robotic intelligence and efficiency.

Computer Software

Responsibilities

Architect hybrid software–hardware systems that implement probabilistic frameworks using energy-based algorithms on thermodynamic chips
Build sampling-based inference systems (e.g., MCMC, Gibbs sampling, variational inference) optimized for thermodynamic computing substrates
Co-design algorithms jointly with hardware teams to map computation efficiently onto novel physical architectures
Deploy, evaluate, and iterate on these systems in real robotic environments
Collaborate closely with physicists, AI researchers, hardware engineers, and product teams to drive real-time adaptive computation
Contribute to publications, patents, and open-source frameworks advancing the field of physical AI and intelligent thermodynamic systems

Qualification

PythonMachine Learning FrameworksProbabilistic InferenceDeep LearningEnergy-based AlgorithmsSystems MindsetInterest in Non-traditional HardwareRobotics Experience

Required

Strong coding ability in Python and at least one ML framework (PyTorch, JAX, or TensorFlow)
Experience with probabilistic inference (MCMC, variational inference, or energy-based models)
Solid understanding of machine learning fundamentals — especially deep learning, Bayesian, and Maximum Entropy Inverse RL
Enthusiasm about both non-traditional hardware (e.g., neuromorphic, analog, quantum, thermodynamic) and how algorithms map to computation beyond GPUs
Interest in developing within the active inference framework
A systems mindset focused on performance, energy efficiency, and robustness

Preferred

Experience with diffusion/score-based models, or generative world models
Interest in control as inference
Robotics experience (simulators or physical robots)

Benefits

Early access to thermodynamic computing hardware.
Collaboration with leading researchers in active inference, generative modeling, and novel computing.
Real robotic platforms for prototyping and deployment.
Remote-friendly culture with periodic on-site collaboration.
Strong support for research, publication, and open-source contributions.
Salary + equity.

Company

Noumenal Labs

twitter
company-logo
Noumenal Labs is a deep tech artificial intelligence company focused on solving fatal flaws in Physical AI and robotics - lack of adaptability, massive retraining costs and poor performance in the real world.

Funding

Current Stage
Early Stage
Company data provided by crunchbase