Software Engineer, Robotics

Deccan AI•San Francisco, CA

About The Position

Deccan AI is hiring a Software Engineer, Robotics & VLA Systems for its Physical AI Team in Mountain View, California. The company builds expert-curated datasets and evaluation infrastructure for frontier AI labs. Its Physical AI practice develops annotation, synthetic data generation, and model evaluation systems for embodied AI. This role builds production software that connects data pipelines, simulation, training loops, and evaluation for Vision-Language-Action (VLA) foundation models. It sits at the intersection of software engineering and machine learning, rather than being a pure research or data engineering role. The engineer will work directly with frontier lab clients and build systems used to train robot foundation models.

Requirements

  • Master’s or Ph.D. in Computer Science, Robotics, or Machine Learning, or equivalent industry experience.
  • At least 2 years of experience shipping production systems, rather than only prototypes.
  • Strong Python and systems skills.
  • Comfort with Linux, Docker, CUDA, and cloud infrastructure such as AWS or GCP.
  • Working knowledge of robot learning, including imitation learning, behavior cloning, diffusion policies, or reinforcement learning.
  • Understanding of how VLA training works, beyond its theory.
  • Experience with at least one robotics simulator: Isaac Sim, MuJoCo, PyBullet, or Gazebo.
  • Experience with robotics data formats or middleware such as ROS/ROS2, URDF/USD, MCAP, or HDF5.
  • Ability to write clean, tested, and documented code.
  • CI/CD and code review experience.

Nice To Haves

  • Hands-on experience with VLA codebases such as RT-2, Octo, OpenVLA, π0, GR00T N1, or LeRobot.
  • Experience with NVIDIA Omniverse, Replicator, or Cosmos for synthetic data.
  • Experience with sim-to-real transfer techniques such as domain randomization, NeRF, or Gaussian Splatting.
  • Experience building ML evaluation frameworks, model benchmarking suites, or RLHF/DPO training pipelines.
  • Open-source contributions in robotics or machine learning.
  • Published research in manipulation, embodied AI, or VLA systems.

Responsibilities

  • Build the VLA training integration layer, including data loaders, format converters, and preprocessing to feed curated real and synthetic datasets into training frameworks such as LeRobot, Octo, OpenVLA, and π0.
  • Develop evaluation and benchmarking infrastructure, including simulation-based rollouts in Isaac Lab, success-rate tracking, regression detection, and automated client reporting.
  • Manage datasets through versioning, schema validation, metadata indexing, and format conversion across Open X-Embodiment, LeRobot HDF5, RLDS, and client-specific formats.
  • Implement sim-to-real transfer tooling, including domain randomization configurations, Cosmos Transfer integration, and quality validation to assess whether synthetic data improves policy performance.
  • Build annotation-platform backend systems for robotics episode data, including task taxonomy APIs, temporal segmentation, quality scoring, and inter-annotator agreement.
  • Integrate with the NVIDIA Isaac ecosystem, including Isaac Sim, Isaac Lab, GR00T-Mimic, and OSMO, to orchestrate synthetic data generation on cloud GPU infrastructure.

Benefits

  • Founding-stage impact and the opportunity to shape the robotics practice and its systems.
  • Work with frontier lab clients, with systems used by teams such as Google DeepMind.
  • Full-stack ownership of systems across annotation backends, VLA training integration, and simulation-based evaluation.
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service