Senior / Staff Technical Program Manager, AI Research

Description

The company builds general-purpose robots that can work in complex, real-world settings and handle rare edge cases. It develops the whole stack: hardware, robot systems, infrastructure, and the foundation world models that control the robots.

The company has raised over $450M and is investing heavily in model research, infrastructure, hardware, and scaling up manufacturing.

Upcoming work covers faster task adaptation, following instructions and demonstrations, higher execution speed, and dexterous manipulation. Progress depends on models, data, hardware, compute, and evaluation becoming ready in the right order.

Success looks like clear handoffs, blockers raised early, faster resolution of resource conflicts, less coordination overhead for technical teams, and leadership that trusts the milestone and risk picture.

Responsibilities

• Own the combined execution plan for research work that spans pre-training, post-training, applied research, data, GPU infrastructure, robot hardware, and delivery of models to customers.
• Link milestones, spot dependencies early, push blockers to closure, and help teams make timely calls on people, compute, and sequencing.
• Give researchers and engineers more uninterrupted build time, and make the path from experiment to usable model easier to forecast.
• Work with research leads, technical leads, and the infrastructure, robotics, and hardware groups. You do not need to design the science yourself, but you need enough depth to follow it, question assumptions, and turn unclear problems into workable options.
• Maintain one current view of experiments, model and data readiness, system integration, releases, dependencies, critical paths, and decision points.
• Coordinate handoffs from pre-training checkpoints into post-training, applied model work, evaluation, and deployment. Agree on handoff criteria, owners, and readiness standards.
• Track when training data, collection systems, robots, workcells, and GPU capacity must be ready. Manage demand forecasts and routine scheduling trade-offs within agreed priorities.
• Find cross-team gaps before they cause delays, get the right people together, keep action plans and fallback options current, and escalate major trade-offs with a specific recommendation.
• Flag early where milestones need more engineering capacity, specialist skills, or re-prioritization. Support leaders' staffing decisions without managing people yourself.
• Run focused program reviews, keep decision and dependency logs, and send short, reliable updates to technical leadership without adding heavy process.
• Separate experiments, proofs of concept, engineering-readiness milestones, and customer deployment commitments. Track confidence and learning goals rather than forcing research into false precision.
• Work with hardware and operations TPMs and program leads to define interfaces, ownership, acceptance criteria, and release expectations as the company grows.

Requirements

• Substantial experience running technically dense programs with interdependent deliverables across software, ML, research infrastructure, robotics, or similar settings, with senior or staff-level judgment.
• Solid grasp of the ML development lifecycle: datasets, training, evaluation, iteration, and the difficulties of putting models into production.
• Proven ability to build an execution system from an unclear start, identify the dependencies that truly matter, and move work forward across teams without formal authority over them.
• Enough technical fluency to engage credibly with researchers and infrastructure engineers, ask pointed questions, and lay out real trade-offs rather than just gather status.
• Experience managing scarce engineering resources such as compute, test systems, labs, hardware platforms, or specialist teams.
• Strong written and spoken communication, able to condense many technical threads into a brief summary leaders can act on.
• Good judgment on what to resolve yourself, when to bring in technical owners, and when leadership needs to decide.
• Background in training large models, foundation or generative model work, computer vision, embodied AI, or robotics.
• Working knowledge of GPU clusters, data pipelines, robot evaluation, or hardware and software integration.
• Past work linking research groups with customer-facing teams in applied engineering or model shipping.

Conditions

The role is at senior or staff level.