ML engineer - API Platform
Physical Intelligence5 hours ago
Responsibilities
- Build Pi’s model API, including data ingestion, fine-tuning, evaluation, remote inference, partner-facing tools, and deployment integrations.
- Design reliable multi-tenant infrastructure that can scale to thousands of organizations and potentially millions of robots.
- Build and operate low-latency inference systems for models controlling robots in real-world environments.
- Develop product workflows from partner data upload and validation through processing, fine-tuning, and evaluation.
- Work with researchers to turn new model capabilities into stable APIs, tools, documentation, and abstractions.
- Write production code integrated with Pi’s infrastructure and define the developer platform for general-purpose robotics.
Requirements
- Strong software engineering fundamentals and experience building production systems.
- Deep backend and systems experience with APIs, services, databases, caching, distributed systems, and infrastructure.
- Experience building and scaling developer platforms, especially platforms for model fine-tuning, inference, or compute-intensive workloads.
- Understanding of reliability, latency, multi-tenancy, versioning, observability, and operational complexity at scale.
- Familiarity with deploying, serving, and debugging machine-learning models in production; ML research experience is not required.
- Strong Python skills and ability to work across infrastructure and product boundaries.
- Experience with low-latency or real-time systems, streaming, inference transport, WebSockets, QUIC, model serving, inference, fine-tuning, or developer-platform infrastructure is beneficial.
- Familiarity with Python, Postgres, ClickHouse, GCP, Kubernetes, Modal, React, and TypeScript is beneficial.
- Experience with security, authentication, authorization, or multi-tenant infrastructure is beneficial.
Tech Stack
About Physical Intelligence
Physical Intelligence builds general-purpose AI for the physical world, developing foundation models and full-stack robotic systems for robots and other actuated devices. The privately held company was founded in 2024 and focuses on taking designs from prototype to high-volume production, including actuation, mechatronics, and manufacturing process development. Its work aims to power today’s robots and future hardware across industries.