Noumenal Labs · 1 month ago
Research Engineer - Probabilistic Perception
United States
Full-time
Remote
Mid, Senior Level
$100K/yr - $150K/yrNoumenal Labs is a deep tech AI company focused on closing performance gaps in outdoor robotics. They are seeking a Research Engineer to develop and deploy probabilistic generative models for perception and spatial reasoning, collaborating with researchers and engineers to build robust systems for real-time applications.
Computer Software
Responsibilities
Develop and deploy probabilistic generative models for perception, scene understanding, and spatial reasoning (structured generative models, inverse graphics, Bayesian scene reconstruction) on hardware in a commercial product
Build inference engines for SLAM, 3D reconstruction, object-centric scene modeling, and spatial world models, leveraging MCMC, variational inference, or novel structured inference techniques
Design systems that combine topological, geometric, and probabilistic methods for robust representation of spatial and conceptual structure
Lead and engage in directed engineering efforts to translate novel algorithms into performant systems suited for real-time or near–real-time perception
Collaborate with researchers in probabilistic computing, robotics, and AI to prototype, test, and iterate on models using synthetic and real sensory data
Qualification
Required
Experience building perception systems in robotics
Ability to translate research concepts into robust, scalable engineering implementations
Strong coding ability in Python and modern ML frameworks (PyTorch, JAX, or TensorFlow)
Expertise in probabilistic inference, structured generative models, or Bayesian approaches (MCMC, variational inference, factorized models, hierarchical generative models)
Experience in 3D perception and spatial AI, including at least one of: SLAM, object-centric modeling, structured scene representations, or probabilistic inverse graphics frameworks
Commitment to open-source contributions and internal cross-lab collaborations
Preferred
Experience with topological data analysis, geometric representations, or mathematical structure in learning systems (e.g., planning in latent spaces)
Strong mathematical background (geometry, topology, optimization, or probabilistic modeling)
Background working in interdisciplinary research groups (AI, neuroscience, robotics, mathematics)
Publications in machine learning, probabilistic modeling, computational neuroscience, or mathematical methods for perception
Benefits
Close collaboration with researchers in robotics, physics-inspired AI, and spatial intelligence.
Access to real-world data for 3D perception and inference experiments.
A remote-friendly environment, flexible work culture, competitive salary + equity.
Company
Noumenal Labs
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.
2-10 employees
Funding
Current Stage
Early StageCompany data provided by crunchbase