About the role
Who we are Lab37 Robotics , is a technology company focused on the development and deployment of robots designed specifically for direct-to-customer food production. Our mission is to revolutionize the food industry by creating innovative robotic solutions that enhance efficiency, quality, and customer satisfaction. We are passionate about pushing the boundaries of technology to deliver cutting-edge products that meet the evolving needs of our clients.
What you'll do Take on ownership of the data and ML platform layer that powers Lab37's robotics fleet. This spans building ETL pipelines that turn raw robot telemetry into actionable analytics, managing the ML training infrastructure that produces the computer vision models running on our robots, and ensuring the observability and reliability of the entire data stack. Work with a small, high-impact platform team to build the systems that every product team at Lab37 consumes — from data scientists training models, to kitchen operations teams viewing dashboards, to engineers deploying new ML models to robots in the field.
Responsibilities Build and maintain ETL pipelines that ingest, validate, transform, and load robot telemetry data into BigQuery for analytics and ML training Manage ML training infrastructure using Argo Workflows on Kubernetes — from data extraction through model training, evaluation, and registration Design and maintain data quality checks and observability dashboards (the "glass panel" for data pipeline health) Own the data warehouse layer — schema design, incremental loading, dbt transforms, and BigQuery/Athena query optimization Build dashboard infrastructure (Superset, Grafana) that kitchen operations and leadership teams rely on for real-time insights Collaborate with data scientists to productionize model training pipelines — from notebook experiments to reproducible, automated workflows Contribute to infrastructure-as-code (Terraform) and CI/CD pipelines for data and ML workloads Participate in on-call rotations for data pipeline reliability