Senior ML Platform Engineer, Robotics

Andromeda Robotics•San Francisco, CA
•Onsite

About The Position

Andromeda is building a social intelligence layer for humanoid robots. Its first robot, Abi, is deployed in aged care and healthcare, and the company is scaling its platform to operate across thousands of real-world environments. This is a hands-on senior individual-contributor role focused on building shared workflows for machine-learning development, including reproducible experiments, evaluations, dependable training data, reliable training infrastructure, and integration with existing robot software. The engineer will implement tools, help teams adopt them, and improve the workflows based on real workloads. The role leads through influence and sets technical direction for ML tooling, without managing a team. It partners with DevOps, cloud, and robot software platform engineers. The technology stack includes AWS, Nix, ROS 2, NVIDIA Jetson, and third-party and custom tools. The position is on-site five days a week in San Francisco.

Requirements

  • At least 5 years of experience in ML infrastructure, developer platforms, and robotics software.
  • Comfort working with robots and embedded systems.
  • Experience building internal tooling adopted across multiple teams and systems that other engineers rely on and find useful.
  • Strong engineering, packaging, and dependency-management skills.
  • Enjoyment of hardening production systems and ability to break complex systems into maintainable, tested components.
  • Confidence working in an unfamiliar codebase.
  • Practical experience with reproducible ML workflows, evaluation, experiment tracking, data lineage, and model artifacts.
  • Ability to judge when lightweight conventions are sufficient and when automation is worthwhile.
  • Experience building ML training-data pipelines, including cleaning, annotation, and dataset quality checks, and training and evaluating classifiers.
  • Experience operating cloud compute and training workloads, with sound judgment about reliability, observability, performance, and cost.
  • Comfort debugging across model code, runtime environments, and robot interfaces, including hands-on hardware testing.
  • Clear communication and collaborative judgment, including advising senior engineers, agreeing ownership with adjacent teams, and supporting colleagues without becoming a bottleneck.

Nice To Haves

  • ML model and code optimization for NVIDIA Jetson and cloud deployments, including quantization, mixed precision, TensorRT or comparable inference compilation, and CPU, RAM, and GPU performance profiling.
  • Experience managing distributed training infrastructure, including multi-GPU and multi-node workloads, scheduling, checkpointing, and failure recovery.
  • Experience integrating ML workloads with ROS 2-based robot software.
  • Experience with Nix/NixOS and AWS.
  • Willingness to learn Nix packaging, store semantics, and binary caches; prior Nix expertise is welcome but not required.
  • Experience supporting simulation or on-robot evaluation systems.

Responsibilities

  • Establish practical patterns for experiment repositories, configuration, code review, and automated checks.
  • Build reusable libraries, scripts, and templates that help engineers start quickly and produce work that others can run.
  • Build multimodal data-lake infrastructure and pipelines to clean, annotate, and serve model-training data.
  • Add dataset quality checks and lineage so datasets can be traced to their source robot data.
  • Make it inexpensive to test ideas against production baselines and challenger models, and establish clear criteria for determining whether a candidate model wins.
  • Define the model progression from candidate through qualification and approval to deployment on the robot.
  • Develop robot integration-testing patterns that test inputs, preprocessing, inference, interfaces, and downstream behavior within the existing robot codebase.
  • Work with engineers on CI, simulation, and hardware-in-the-loop testing to catch failures software-only testing may miss.
  • Own ML-specific workflows for provisioning compute and scheduling and managing training jobs.
  • Work with cloud and DevOps engineers on underlying infrastructure, access, reliability, and cost.
  • Build reliable model packages and dependency configurations, and learn and apply Nix packaging, the Nix store, and binary caches.
  • Contribute to robot build and release workflows so models reach Jetson with compatible dependencies, tested runtime behavior, and a rollback path.
  • Support staged rollouts or shadow deployments and monitor deployed model behavior on robots in the field.
  • Build dashboards connecting model and dataset versions, evaluation results, and deployed performance to agreed product metrics.
  • Help teams distinguish benchmark gains from improvements in Abi's behavior, reliability, and user experience.
  • Work with engineers to understand bottlenecks, advise on technical choices, and make shared tools useful in practice.
  • Provide clear examples, documentation, and handoffs.

Benefits

  • Benefits
  • Equity
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service