P
Physical Intelligence
General-purpose AI foundation models for robots
About Physical Intelligence
Physical Intelligence (often shortened to Pi) is a San Francisco-based artificial intelligence company developing general-purpose foundation models that let robots and other physical machines carry out a wide range of real-world tasks. Founded in 2024 by a group of researchers and engineers from Google DeepMind, Stanford, UC Berkeley and the robotics industry, including Karol Hausman, Sergey Levine, Chelsea Finn, Brian Ichter, Lachy Groom and Adnan Esmail, the company aims to build a single learned "brain" that can control many different types of robots.
The company's main research output is its pi series of vision-language-action models. The first model, pi0, combined a pretrained vision-language model with a method for generating continuous motor actions, and was trained on data from multiple robot platforms performing tasks such as folding laundry, clearing tables and assembling boxes. Later versions, including pi0.5 and subsequent releases, focus on generalising to new homes and environments the system has never seen, handling longer multi-step tasks and learning from experience.
Rather than building its own robot hardware, Physical Intelligence works across a variety of arms, mobile manipulators and other platforms, collecting large and diverse datasets of robot experience and pairing them with web-scale data. It has released some model weights and code as open source for researchers, alongside publishing research on robot learning, reinforcement learning and model training.
The company's long-term goal is general-purpose physical intelligence that can be applied across household, industrial, logistics and other settings, with partners and robot makers using its models as a base for their own applications.
Source: Dealroom company dataOpen roles
| Role | Team | Location | Posted |
|---|---|---|---|
| ML Infra Engineer, ComputeData & AI | Data & AI | San Francisco | 6d ago |
| Robot Service Technician Shift LeadOperations | Operations | San Francisco | 1w ago |
| IT Operations EngineerEngineering | Engineering | San Francisco | 1w ago |
| 3D Printer TechnicianOperations | Operations | San Francisco | 1w ago |
| Fleet Inventory SpecialistOperations | Operations | San Francisco | 1w ago |
| Deployment & Engagement Operations LeadOperations | Operations | San Francisco | 2w ago |
| Lead Security EngineerEngineering | Engineering | San Francisco | 1mo ago |
| Environment SpecialistOperations | Operations | San Francisco | 1mo ago |
| Environment CoordinatorOperations | Operations | San Francisco | 1mo ago |
| General Manager, Research OperationsOperations | Operations | San Francisco | 1mo ago |
| Learning & Development Lead, OperationsOperations | Operations | San Francisco | 1mo ago |
| Recruiter, HardwarePeople | People | San Francisco | 1mo ago |
Physical Intelligence in the San Francisco ecosystem
The Physical Intelligence office with the hot startups, most funded companies and investors around it. Hover or tap a pin.
Dealroom maps whole ecosystems like this one for cities, regions and governments. See the ecosystem platform →
Investors
From Dealroom's funding and investor records, grouped by the round each investor first entered.
Seed6 investors entered at this stage
Thrive Capital
Sequoia
Anti Fund
Lux Capital
Khosla Ventures
OpenAI Startup Fund
Series A2 investors entered at this stage
Bezos Expeditions
Jeff Bezos
Series B and later5 investors entered at this stage
Source: Dealroom Talent Intelligence.
See Physical Intelligence on the full Dealroom platform
Live deal flow, founder pedigree, hiring signals, revenue trajectory, and the full cap table.
Nidus raises $30M seed to build defense robotics AI in New York
Machenta raises $40M seed to train robots without real-world data





