P

Physical Intelligence

Develops AI-driven robotic systems for industry

NPI Technical Program Manager

Full-Time
No salary listed
Mid
Bachelor's
San Francisco, CA, USA
In PersonDomestic and occasional international travel to vendors and contract manufacturers is required.
Project & Program ManagementProgram Management

About the job

Requirements
  • A Bachelor's degree in Engineering or a related field.
  • Experience shipping hardware.
  • At least 3 years of experience in operations, supply chain, manufacturing program support, or engineering program management, including cross-functional and vendor engagement.
  • Strong organizational and project management skills, with the ability to drive process improvements and manage multiple priorities in a fast-paced environment.
  • Proficiency with Notion, Linear, Excel, project trackers, and procurement, ticketing, or workflow systems.
  • Ability to translate engineering detail into clear status updates for technical and executive audiences.
  • Willingness to travel domestically and occasionally internationally to vendors and contract manufacturers.
  • Familiarity with R&D operations, robotics, or other technical environments involving complex procurement, NPI, or capitalization requirements.
  • Familiarity with electromechanical systems, PCBA, machined parts, or robotics hardware.
  • Direct experience with build planning, bill of materials management, Clear-to-Build analysis, or material readiness for hardware programs.
  • Willingness to spend time on the build floor working alongside technicians and engineers.
  • Ability to identify and implement process improvements using data and cross-functional input.
  • Familiarity with enterprise resource planning, materials requirements planning, product lifecycle management, or production tracking tools.
Responsibilities
  • Partner with Platform leads and the engineering team to drive program schedules, timelines, and end-to-end execution.
  • Own build planning and production coordination for hardware platforms from part and tooling kickoff through production output to fleet hand-off.
  • Drive material forecasting, Clear-to-Build analysis, dashboards, and readiness reporting to keep build schedules accurate and risks visible.
  • Manage hardware staging and acceptance hand-offs to the fleet integration team.
  • Coordinate the arrival of commercial off-the-shelf hardware from global vendors, including modifications for repair, allocation, tracking, replenishment, and end-of-life management.
  • Run logistics readiness and troubleshooting for incoming components.
  • Route custom hardware requests across internal build and fabrication spaces and external vendors.
  • Partner with supply chain and procurement on vendor follow-up, lead-time management, and risk mitigation for long-lead and single-source parts.
  • Identify and implement process improvements in how the company plans, builds, and ships hardware as the program scales.

About the company

P

Physical Intelligence

View
Website

Physical Intelligence develops AI-powered robotic systems by building general-purpose AI models and learning algorithms to control physically actuated devices in the real world. Its product is intelligent control software that runs on robots, using foundational models to sense, decide, and act in real environments, often integrating with hardware and partner ecosystems. The company differentiates itself by focusing on foundational AI for physical robotics and drawing support from top venture investors to fund both R&D and commercialization, aiming to address multiple industries. Its goal is to bring AI-driven robotic systems to manufacturing, healthcare, and logistics, improving efficiency and productivity through learnable robotics and broad deployment.

Company Size

201-500

Company Stage

Late Stage VC

Total Funding

$2.1B

Headquarters

San Francisco, California

Founded

2024

Get referred to Physical Intelligence

See people who can refer or advise you

Simplify Jobs

Simplify's Take

What believers are saying

  • March and September 2026 leases add 374,500 square feet, signaling aggressive scaling.
  • May 2026 funding lifted total capital above $2 billion, extending runway for model development.
  • π0.7 matched specialist systems on folding, coffee, and box assembly, strengthening product credibility.

What critics are saying

  • π0.7 still fails complex multi-step commands; full autonomy remains unsolved in 2026.
  • Competition from OpenAI, Google, NVIDIA, and Anduril will compress Physical Intelligence's edge fast.
  • A $11 billion valuation without revenue makes the company existentially dependent on next funding rounds.

What makes Physical Intelligence unique

  • April 16, 2026 π0.7 generalizes across robots, tasks, and new language instructions.
  • Founders include Sergey Levine, Chelsea Finn, and Karol Hausman; elite robotics pedigree matters.
  • Pi website shows FAST, MEM, and RL token systems for long-horizon embodied control.

Help us improve and share your feedback! Did you find this helpful?

Benefits

  • Health Insurance
  • Dental Insurance
  • Vision Insurance
  • Unlimited Paid Time Off
  • Remote Work Options
  • Hybrid Work Options
  • Stock Options
  • Company Equity
  • 401(k) Retirement Plan
  • Wellness Program
  • Mental Health Support
  • Phone/Internet Stipend

Growth & Insights and Company News

Headcount

6 month growth

-6%

1 year growth

1%

2 year growth

1%
Unite.AI
Oct 13th, 2026
Physical Intelligence opens Pi API beta for robot foundation models.

Physical Intelligence opens Pi API beta for robot foundation models. Published October 13, 2026 Orion Sato, Robotics & Automation AI Research Agent Physical Intelligence announced the Pi API beta on October 13, 2026, making its latest robotic foundation models available to outside developers for hosted fine-tuning and inference. Access to the beta runs through a waitlist, with applicants asked to describe what they want to build. Under the workflow laid out in the announcement, a developer uploads robot data, launches fine-tuning on compute and infrastructure that Physical Intelligence manages, and accesses the fine-tuned model through a hosted gateway to run inference on their robot. Evaluation data then guides the next iteration. The company said the latest model available through the API already delivers more data-efficient fine-tuning and improved capability for its partners, and that because it can update the models underpinning the API rapidly, it plans to continue enabling access to its latest and most capable models. Physical Intelligence described the launch as its first step toward providing a physical intelligence foundation the community can build upon and enabling end-to-end autonomy. The models behind the API. The announcement frames general-purpose robotic foundation models such as π0 and π0.5 as a promising way to tackle the challenge of deploying robots: beyond identifying the right use case and building the right hardware, developers need models intelligent enough to perform the task at hand. These models provide a strong foundation of physical knowledge and capabilities that substantially reduces the data and effort needed to adapt to new tasks, environments, or robots, the company said, but adapting them still requires ML infrastructure, compute, and complex data pipelines. Physical Intelligence publishes open-source versions of these models in its openpi repository. The repository contains three types of models: π0, a flow-based vision-language-action model; π0-FAST, an autoregressive vision-language-action model based on the FAST action tokenizer; and π0.5, an upgraded version of π0 with better open-world generalization, trained with a method the company calls knowledge insulation. Base checkpoints for all three are pre-trained on more than 10,000 hours of robot data, and the repository is released under an Apache-2.0 license. Physical Intelligence released π0.5 and PyTorch support for the repository in September 2025, after adding instructions for training vision-language-action models on the full DROID dataset in June 2025. Running the open models locally requires an NVIDIA GPU with more than 8 GB of memory for inference, with the RTX 4090 given as the example, and more than 70 GB of memory on hardware such as the A100 or H100 for full fine-tuning. The repository has been tested on Ubuntu 22.04 and does not currently support other operating systems. It also supports remote inference, in which a model runs on a separate server and streams actions to the robot over a websocket connection. The repository notes that π0 was developed for the company's own robots, which differ from widely used platforms such as ALOHA and DROID, and that the team does not expect every adaptation attempt to succeed. The API takes that pipeline over as a hosted service: users supply the data, and Physical Intelligence manages the training compute and serves the finished model through its gateway. Partner deployments across industries. Physical Intelligence said it sought to answer two questions with the API: whether its models support a broad range of applications and robotic hardware in diverse industries, and how easily the API enables developers to iterate toward useful, practical autonomy. To test breadth, the company works closely with companies deploying robots, using these collaborations to validate how well its models perform across different tasks, hardware, and operating environments. The announcement features Actor, a company located in Utah that deploys autonomous heavy machinery. In footage the company described as a real commercial deployment, Actor's autonomous wheel loader scoops material and loads it into a dump truck without any operator inside the machine. Other collaborations named in the announcement include AgRobotics, CloudChef, a dairy partner, Ultra, and Weave. Across these initial collaborations, the company said, a common pattern emerged. The challenges differed from task to task, from an inside-out garment to an ingredient that would not stay in a scoop to a manufacturing part that landed in the wrong position, but the improvement process was much the same: collect demonstrations, train a model, test it on the robot, and use the failures to decide what data to collect next. Physical Intelligence said it built the API to let developers run this iteration loop themselves, using its latest models. Early-Access results and next steps. To test independent iteration, Physical Intelligence gave a small number of companies early access to the API and said it was impressed by how quickly they got started and began making progress. The announcement highlights Almond, based in San Francisco, California, which builds Axol, a dual-arm robot for physical AI builders whose stated goal is for Axol to build itself. According to the company, Axol deburred spacers used in its own assembly while running a model the team obtained after iterating with the Pi API for four days. Other early-access users shown include Argon Robotics, Chef Robotics, Deft, Dream Machines, Formic, Foundry, Frontier Machines, General Instinct, Intuition.dev, Omakase, robot.com, Ultra, and Yondu, alongside a hackathon project. Most of the applications shown ran with cloud-hosted inference through the API, spanning deployments from Japan to Switzerland to sites across the United States. The company said these early deployments covered a range of environments and tasks and highlighted areas for improvement, including network connectivity challenges for some. Physical Intelligence said the examples show the models it develops can be used for real, useful applications and that its partners are able to use the API to independently iterate on policies and improve autonomy on their tasks, while characterizing the effort as still early, with feedback from developers set to help shape future releases. The company said it is hiring across research, infrastructure, hardware, people, and operations. Orion Sato is an AI-generated research agent focused on robotics, automation, and intelligent machines. His writing explores how advances in robotics are reshaping manufacturing, logistics, healthcare, and everyday life through increasingly autonomous systems. With a technical and execution-oriented perspective, Orion analyzes robotic architectures, sensor fusion, control systems, and the convergence of AI with mechanical intelligence. He is particularly interested in how automation moves from controlled environments into real-world deployment, where reliability, safety, and efficiency matter most. Articles authored by Orion Sato are AI-generated and reviewed by Unite.AI's editorial team to ensure technical accuracy, clarity, and compliance with editorial standards.

Crypto Briefing
Oct 9th, 2026
Ultra raises $62M and deepens ties with robot-brain startup Physical Intelligence

Ultra raises $62M and deepens ties with robot-brain startup Physical Intelligence. The Brooklyn robotics startup rents warehouse robots by the month, and investors just put $62 million behind that model an hour ago Sponsored: Vera - AI-powered prediction market intelligence, built for serious analysts Explore Vera Ultra, a Brooklyn robotics startup that rents robots to warehouses the way you might rent a forklift, has raised $62 million. It also expanded its partnership with Physical Intelligence, the AI company building software that serves as a robot's brain. The $62 million came in two parts. The bigger piece is a $50 million Series A led by Framework Ventures, with Y Combinator also participating. The rest is an earlier $12 million seed round. Y Combinator and Next View led that one. Ultra was founded in 2024, which makes this a fast climb. Going from founding to a $50 million Series A in about two years is quick even by startup standards. The company builds dual-arm stationary robots called Operator, or OP1. They handle packing, sorting, and kitting tasks inside e-commerce warehouses. Kitting, for the uninitiated, means bundling several items into a single package. Think of a subscription box that needs a shampoo, a conditioner, and a sample serum dropped in together. Ultra says its robots have packed over 500,000 orders at sites across the US. One notable deployment sits at Highline Commerce in Brooklyn, where the robots handle up to 30% of the company's fulfillment volume. Setup is also built for speed. Ultra's robots are designed to be installed in a matter of hours, not the weeks or months a full automation overhaul can demand. AI, tech, and the markets they move - in one daily briefing. Daily. Free. Join 34,000+ readers across crypto, finance, and policy. Where Physical Intelligence fits in. Ultra makes the bodies. Physical Intelligence, often called Pi, supplies the brains. Pi is a San Francisco company, also founded in 2024. It builds what are called vision-language-action models, or VLAs. Pi's π0.6 model reportedly hit 96.4% autonomy during full shifts in real-world warehouse settings. That figure comes from deployment data released in February 2026. Pi is not a small player in this space. Its valuation reached $5.6 billion after a $600 million Series B in November 2025. So the partnership pairs a young hardware company with one of the better-funded AI software firms in robotics. Ultra gets a capable brain without building one from scratch. Why the rental model matters. Ultra's robots-as-a-service approach, sometimes shortened to RaaS, lowers the upfront cost for warehouse operators, who might otherwise balk at buying machines outright. The model also gives Ultra room to raise prices as it proves its value. What this means for robotics and its backers. The deal highlights a split that is becoming common in robotics: software and hardware developed by different companies. Pi focuses on the general-purpose brain, while Ultra focuses on practical hardware for a narrow, well-defined job. The flip side is that improvements to Pi's models could, in principle, make Ultra's existing fleet better without new hardware. For investors, the round signals continued appetite for robotics companies that can show revenue, not just research. Ultra's 500,000-plus orders and its Highline Commerce deployment give backers something concrete to point to. Disclosure: This article was edited by Diego Almada Lopez. For more information on how Cryptobriefing create and review content, see its Editorial Policy.

The Real Deal
Sep 30th, 2026
Physical Intelligence takes 80,000 sq ft from Airbnb after $1B raise

Physical Intelligence has agreed to sublease 80,000 square feet from Airbnb at 850 Brannan Street in San Francisco's Showplace Square. The robotics startup is rapidly expanding following a $1 billion funding round in May, which brought its total funding above $2 billion. The deal follows Physical Intelligence's 60,000-square-foot lease at 808 Brannan in March and a recent 233,500-square-foot lease in Mountain View. The company, founded in 2024, is developing general-purpose AI for robots that can perform physical tasks like folding laundry without task-specific training. The expansion reflects growing investor interest in robotics, with venture capital pouring more than $12 billion into robotics startups last year. Physical Intelligence's backers include Thrive Capital, Sequoia Capital and OpenAI Startup Fund.

AsumeTech
Sep 12th, 2026
Why sequoia is betting on Mecka AI to scale action-state data for robotics.

Why sequoia is betting on Mecka AI to scale action-state data for robotics. Modified date: September 12, 2026 Discover more Casual Games The intensifying venture capital race to solve the "data bottleneck" highlights a challenge currently hindering the development of humanoid robots and autonomous systems. Mecka AI, a Palo Alto-based startup, has emerged as a key player in the "Physical AI" sector. Unlike Large Language Models (LLMs) that were trained on vast archives of internet text and images, robotic systems require "action-state" data - precise records of how a physical machine moves and reacts to its environment. This type of data cannot be easily scraped from the web, leading to a scarcity that has made specialized data providers like Mecka AI highly attractive to top-tier investors. In early 2024, Mecka AI announced $60 million in Series A funding to expand its infrastructure for powering physical intelligence. The demand for high-quality robotic training sets is outpacing even the most aggressive growth projections for the sector. The Shift from LLMs to Physical AI. Industry analysts note that while the first wave of the AI boom focused on generative text and coding, the current frontier is moving toward models that can navigate and interact with the real world. This transition has hit a "data wall." While an AI can learn to write a poem by reading millions of books, it cannot learn to pick up a fragile object or navigate a cluttered warehouse without high-fidelity sensor data, teleoperation records, or sophisticated synthetic simulations. Mecka AI competes in an increasingly crowded field that includes startups like Physical Intelligence (Pi) and established players looking to standardize how robots learn. The technical challenge remains immense: collecting real-world data is slow and expensive, often requiring human operators to guide robots through tasks thousands of times. Companies in this space are currently experimenting with hybrid approaches, combining real-world sensor data with "synthetic data" generated in simulated physics environments to accelerate the training process. As humanoid robot prototypes from companies like Figure, Tesla, and Boston Dynamics move closer to commercial deployment, the value of the underlying data used to make them functional continues to drive significant interest in the startups providing it. Discover more Word Games Puzzles & Brainteasers TV & Video

The Tifton Gazette
Aug 31st, 2026
AI is reshaping the Unmanned Ground vehicle Market - mega funding, defense contracts, and edge intelligence drive a new era of autonomous systems.

AI is reshaping the Unmanned Ground vehicle Market - mega funding, defense contracts, and edge intelligence drive a new era of autonomous systems. GlobeNewswire | BCC Research LLC Today at 1:22am PDT Boston, Aug. 31, 2026 (GLOBE NEWSWIRE) - Unmanned ground vehicles are no longer an emerging concept - they are an active investment category commanding billion-dollar funding rounds, sovereign defense contracts, and industrial-scale deployments across mining, agriculture, and logistics. BCC Research's newly published AI Impact on Unmanned Ground Vehicles (UGVs) Market - BCC Pulse Report examines how artificial intelligence is restructuring the competitive dynamics, technology stack, and capital flows of the global UGV market, providing investors with a granular assessment of where value is being created and where risk remains elevated. Key Findings - Massive capital concentration signals market inflection: The UGV sector is entering a rapid expansion phase driven by a software-first investment strategy, with funding concentrated in megadeals for companies capable of mass-producing AI software and hardware at scale. Anduril Industries raised $5 billion in May 2026 - expected to double its valuation to approximately $60 billion - while Physical Intelligence reached a $5.6 billion valuation following a $600 million funding round led by Alphabet's CapitalG, with participation from Jeff Bezos and Thrive Capital. - Government spending is creating guaranteed demand: The U.S. government approved an $839 billion defense bill in February 2026 with a heavy focus on AI, space security, and autonomous systems. This legislative commitment de-risks procurement for manufacturers, converting pipeline into contracted revenue. Ondas Holdings secured a $30 million demining contract in Israel shortly after the bill's passage, while the Emirati defense group EDGE secured a contract supplying the UAE military with 60 advanced robotic vehicles, including 20 armed combat units and 40 THeMIS unmanned ground units. - Labor shortages and personnel safety are accelerating commercial adoption: Persistent labor shortages across agriculture, mining, logistics, and construction - amplified by aging populations particularly across Asia-Pacific - are driving economic pressure to automate. Vale planned to expand its autonomous truck fleet from 32 to 150 vehicles within two years, reporting an 11% productivity increase from self-driving operations. China's Huaneng Group launched the largest fleet of autonomous electric mining trucks at the Yimin open-pit coal mine, featuring 100 driverless vehicles with a 20% improvement in transport efficiency. - Edge-native AI is unlocking previously inaccessible deployment environments: Advances in edge computing now allow UGVs to process data and make autonomous decisions locally - without cloud connectivity or GPS - enabling deployment in contested battlefields, underground mines, and remote industrial sites. This capability is foundational to both defense and commercial use cases and is driving investment in multimodal sensor fusion combining LiDAR, camera, radar, and GPS technologies. - Modular systems and retrofit capability are expanding the addressable market: The ability to upgrade existing fleets with AI capabilities rather than replacing entire vehicles dramatically lowers the capital barrier to adoption. This modular approach, formalized through Modular Open Systems Architecture (MOSA) frameworks, expands the market to organizations with large legacy fleets, accelerating near-term penetration without requiring greenfield infrastructure investment. - The competitive landscape spans defense primes, industrial giants, and AI-native challengers: Key players include Anduril Industries, General Dynamics Land Systems, Milrem Robotics, Textron Systems, Helsing, Shield AI, Carnegie Robotics, Forterra, Hyundai Motor Group, Caterpillar, Hanwha, Hyundai Rotem, Inceptio Technology, Monarch Tractor, Carbon Robotics, Starship Technologies, Physical Intelligence, Advanced Machine Intelligence (AMI), SafeAI, Kodiak, Pronto, Mind Robotics, Robotic Assistance Devices (RAD), Ondas Holdings, VisionWave Holdings, Rio Tinto, Vale, Huaneng Group, John Deere, and EDGE. Strategic Implications The structural forces reshaping this market are mutually reinforcing. Defense budgets are funding large-scale procurement while simultaneously validating the technology for commercial operators, compressing the adoption timeline across sectors. The software-first investment thesis - where market value migrates from physical hardware to intelligent software capabilities - is concentrating capital in companies that can deliver machine learning, computer vision, and autonomous decision-making at industrial scale. Hyundai Motor Group's commitment to increase investment to $26 billion in robotics and intelligent systems from 2025 to 2028, alongside Caterpillar's pledge to more than double its technology investments by 2030, signals that legacy industrial platforms are repositioning aggressively rather than ceding ground to AI-native entrants. Regulatory complexity, however, introduces meaningful friction. The EU AI Act mandates that autonomous systems be explainable, ethical, and auditable - requirements that compel manufacturers to engineer White-Box AI architectures at significant additional cost and development time. Simultaneously, interoperability challenges across multi-vendor fleets and the computational constraints of edge AI hardware remain active barriers to faster market penetration, particularly for smaller commercial operators who cannot absorb the capital intensity of full-fleet deployment. Investment Considerations The UGV sector presents a high-conviction growth thesis anchored in legislative spending mandates, demographic-driven labor economics, and a secular shift toward AI-defined hardware. Near-term upside is most visible in defense-oriented players with contracted government revenue - Forterra's $238 million Series C, secured following multiple Department of Defense contracts, exemplifies this de-risked profile. In commercial sectors, agricultural robotics firms such as Monarch Tractor and Carbon Robotics, and logistics automation providers including Inceptio Technology and Starship Technologies, are capturing the labor substitution opportunity at scale. Risks are concentrated in regulatory compliance costs, the capital intensity of infrastructure deployment, and the technical challenge of reliable autonomous performance in GPS-denied environments. Investors with longer time horizons should monitor companies building proprietary edge AI capabilities and sensor fusion stacks, as these assets are likely to define competitive differentiation as hardware commoditizes. About the Report The AI Impact on Unmanned Ground Vehicles (UGVs) Market - BCC Pulse Report provides qualitative analysis of AI-driven disruption across the UGV market, covering technology trends, investment activity, use case analysis, key player profiling, emerging technologies, and strategic challenges shaping the sector's trajectory. About BCC Research BCC Research provides objective, unbiased measurement and assessment of market opportunities with detailed market research reports. Our experienced industry analysts assess growth trends, identify and evaluate new and changing market opportunities, and provide critical information and innovative decision support tools to help inform the strategic decision-making process. For media inquiries, email [email protected] or visit our media page for access to our market research library. Any data and analysis extracted from this press release must be accompanied by a statement identifying BCC Research LLC as the source and publisher. BCC Research LLC 50 Milk St., Ste. 16, Boston, MA 02109 [email protected] | +1 781-489-7301 www.bccresearch.com This is a paid placement. For further inquiries, please contact GlobeNewswire directly.