MLOPs Engineer

Everything humanity depends on is mined, dug, or grown. At AIM, we are building the autonomous linchpin of civilization. We transform heavy machinery—bulldozers, loaders, excavators—into AI-powered fleets that operate continuously, safely, and at peak performance in the world’s harshest environments.

AIM runs production mines, large scale infrastructure builds, and defense operations as a TRL9 hardened system, not a science experiment.

Built by engineers from mining, construction, Waymo, SpaceX, Google and Tesla, AIM enables scalable earthmoving, turbocharging the global economy’s physical foundation. AIM is backed by some of the most sophisticated capital in the world, including General Catalyst, Khosla Ventures, Elad Gil, Human Capital, Ironspring Ventures, Mantis, DCVC.

• Design, build, and maintain automated CI/CD/CT pipelines for Machine Learning models. Design the deployment workflow to validate model safety before physical deployment.

• Optimize Machine learning models by building inference engines. You will manage GPU resource allocation on embedded devices to balance latency with power constraints.

• Provision and manage scalable cloud infrastructure to support distributed training clusters and simulation environments.

• Design and maintain a multi-modal data pipeline capable of ingesting and synchronizing massive streams of LiDAR, Camera, and IMU data. You will build the storage and retrieval layers that fuel our training sets.

• Implement robust monitoring to track model performance and data drift. You will be responsible for building telemetry to monitor system health and inference latency on the edge to ensure robot safety.

• Implement Feature Stores and strict Data Versioning to ensure total reproducibility of models. You will ensure we can trace every robot decision back to the exact code, data, and model version used.

• Maintain and evolve AIM’s ML data platform, including large-scale multi-modal data pipelines (LiDAR/Camera/IMU), schema/metadata standards, dataset indexing, and high-throughput storage + retrieval layers to support training, evaluation, and simulation workflows.

• 3+ years of professional experience in MLOps, DevOps, or Software Engineering, with a focus on productionizing ML models.

• Proficiency in Python and C++

• Proven experience optimizing Deep Learning models with CUDA ONNX and TensorRT

• Experience building large-scale ML data lakes and dataset pipelines (e.g., Ray Datasets / Ray Data) for ingestion, transformation, versioning, and retrieval of multi-modal training data.

• Strong command of Docker and Kubernetes

• Hands-on experience with cloud architecture.

• Mastery of CI/CD tools

• Familiarity with handling non-structured robotics data formats (e.g. Rosbags, MCAP, PCD files)

• Experience with high-throughput data tools like Apache Kafka, Airflow, or Spark

Ideal candidate will have additional experience in several of the following areas:

• Experience working with ROS / ROS2

• Experience deploying to embedded compute platforms

• Experience integrating ML pipelines with simulators

Joining AIM, You Will

• Solve a massive set of real-world problems that require scalable earth moving

• Accomplish that via deploying and expanding cutting edge tech

• Run your workstreams with the largest degree of autonomy

• Have an opportunity for rapid growth and a large voice on the direction of the company

• Receive company funded medical, dental, vision, 401k, life insurance, gourmet food & perks

• Experience a strong onsite collaboration (AIM offices, labs and proving grounds on the east side of the Greater Seattle area)

• Have an opportunity to travel to unique sites around the world (Americas, Australia, Africa & more)