SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Dyna Robotics is hiring a Product Manager for its Data Engine, a core platform that transforms field robot data into training signals for continuous model improvement. This role owns the complete data pipeline—from defining what robots capture in the field, to designing labeling schemas, to ensuring quality standards across annotation workflows. You'll also own the internal platform tooling used by research, engineering, and operations teams for labeling, visualization, and model evaluation, as well as the operator-facing interfaces that field teams and customers use to run and monitor robots at deployment sites.
Key responsibilities include: defining instrumentation and logging standards across the robot fleet; owning the labeling schema and auditing processes to ensure it stays aligned with real-world conditions; setting and maintaining quality bars for labeled data, deciding between in-house and vendor annotation work; managing the full loop of episode capture, annotation pipelines, model-behavior observability, and remote teleoperation for recovery data; designing internal tools for technical users who demand speed and direct access; and creating customer-facing dashboards that build trust in deployed systems.
You'll spend significant time on customer sites observing robots in operation, understanding failure modes, and learning what operators and customers actually need—requirements that often go unfiled as tickets. The role requires balancing competing constraints: coverage vs. quality in labeling, speed vs. guardrails in internal tools, and technical depth vs. accessibility for non-technical field operators.
The ideal candidate brings 4+ years of product management experience owning data pipelines for ML systems end-to-end, familiarity with ML infrastructure and training pipelines, and experience designing tools for both technical and non-technical users in real operational environments. Bonus experience includes video/multimodal data labeling, robotics or autonomous vehicle backgrounds, human-in-the-loop systems, deployed hardware in customer sites, and multi-site fleet operations.
