AI Productivity Tools Engineer

Physical Intelligence; San Francisco (CA); On-site; USD 140,000 - 230,000; Full time

Job summary

Physical Intelligence is hiring a Software Engineer focused on AI productivity to build and roll out internal tools that accelerate AI use across the company. You will collaborate with engineering, research, operations, and people teams to understand workflows and create reliable tools that fit how they work.

You will build backend services, user interfaces, and integrations, drive adoption, and help scale AI solutions across PI with a focus on speed, reliability, and safe access to data.

Qualifications

• Strong software engineering fundamentals and the ability to ship quickly.
• Hands-on fluency with AI coding workflows and modern LLM-based tools.
• Experience building internal tooling and developer platforms.
• Good judgment around security, permissions, and data access.

Responsibilities

• Own AI tooling adoption across the company and drive user adoption.
• Build internal AI tooling, integrations, and agent infrastructure.
• Create ergonomic AI agents and workflows that are easy to use and monitor.
• Develop tools to increase engineering, research, and operational velocity.
• Share playbooks, onboarding, and demos to spread best practices.
• Collaborate with security and data teams to ensure proper access controls.

Skills

AI tooling

Backend services

LLM integrations

Developer experience

Security & data access

Tools

React

TypeScript

Python

Postgres

ClickHouse

GCP

Kubernetes

Job description

Physical Intelligence is hiring a Software Engineer focused on AI productivity to build and roll out internal tools that accelerate AI use across the company. You will collaborate with engineering, research, operations, and people teams to understand workflows and create reliable tools that fit how they work.

You will build backend services, user interfaces, and integrations, drive adoption, and help scale AI solutions across PI with a focus on speed, reliability, and safe access to data.