Technical Program Manager, Datacenter Optimization

Overview

Physical Superintelligence is an exceptionally well-funded seed-stage startup with roots at Google, Meta, DeepMind, NVIDIA, Citadel, Harvard, MIT, Johns Hopkins, the Perimeter Institute, and the Institute for Advanced Study building AI systems to discover new physics at scale. We are seeking a Technical Project Manager to own the end-to-end delivery of our datacenter optimization engagements — from first technical scoping call through deployed, verified results in a live facility.

Our mission is to discover and commercialize transformative physics breakthroughs at scale with artificial superintelligence, safely, verifiably, and for broad public benefit.

The last century's golden age of physics gave us transistors, lasers, and nuclear energy. We believe artificial superintelligence will unlock the next one. We're creating the infrastructure to industrialize scientific discovery and usher in this new era.

Datacenters are where that discovery meets the physical world first. Thermal transport, fluid dynamics, power delivery, and materials behavior all compound into the efficiency of a facility, and all of them are places where better physics translates directly into megawatts, dollars, and carbon. This role is how that work reaches customers.

Role

You own the customer and you own the project. Our optimization engagements sit at the intersection of a research team producing novel physical models and an operator running critical infrastructure with real uptime obligations. You are the person accountable for translating between them and for the outcome on both sides. This role is deliberately scoped for someone who wants more surface area than a typical TPM job offers, and it concentrates on four areas.

Responsibilities

Own engagements end to end. Run datacenter optimization projects from technical discovery through deployment and verified results. Define scope, success criteria, and measurement methodology with the customer up front. Build and hold the plan, the critical path, the risk register, and the change control. When a project is at risk, you know before the customer does, and you have a plan when you tell them.

Own the customer relationship. Be the primary technical point of contact for facility engineers, infrastructure leads, and their executives. Run the working cadence, the steering reviews, and the hard conversations about scope, timeline, and results. Earn enough credibility that customers bring you problems early instead of escalating late. You will spend real time on-site in mechanical rooms and on whitespace floors, not only in conference rooms.

Bridge research and deployment. Take physics and modeling output from our research team and drive it into something that can be validated against a real facility, instrumented, and operated safely. Own the interface: the data we need from the customer, the constraints they cannot violate, the instrumentation and telemetry plan, the baselining approach, and the verification standard we hold ourselves to. Push back on both sides when the science and the operating reality are not yet reconciled.

Run delivery AI-first. This is non-negotiable: you are, or are willing to become, completely AI-native, fluent across the current frontier models and the best existing agent and automation tooling, and you keep pace as they change. Project management, status synthesis, data reconciliation, reporting, and customer artifacts should be substantially agent-driven. You are expected to build the delivery machinery that lets one TPM carry a portfolio that would normally take a team, and to turn each engagement into reusable process rather than bespoke heroics.

Across these areas you also shape what our commercial delivery function becomes. You are early enough that the playbook does not exist yet. What you build — the scoping framework, the measurement standard, the delivery model — becomes how PSI ships to infrastructure customers.

What We're Looking For

A proven owner of complex technical projects. Five or more years running technical programs where you were genuinely accountable for the outcome, not just the tracker. You have delivered projects with hardware, facilities, or physical-world dependencies, where a slip has consequences you cannot recover with a hotfix.

Datacenter or critical infrastructure fluency. You understand how a datacenter actually works: power chain, cooling topology, airflow and thermal management, PUE and its limits as a metric, BMS and DCIM systems, capacity and commissioning. You can hold a technical conversation with a facility engineer and be taken seriously. Direct experience in colocation, hyperscale, HPC, or mission-critical facilities is strongly preferred.

Customer ownership at the technical level. You have been the face of a technical delivery to a demanding customer and kept the relationship strong through bad news. You can negotiate scope, set expectations that hold, and represent your team's work credibly to an executive buyer.

Agentic, hands-on. You are ready to build, orchestrate, and operate multi-agent workflows. You are familiar with current agent frameworks and you think in terms of what to automate and how to compose agents into reliable workflows.

An instinct to instrument. You are comfortable with data — SQL, BI tooling, time-series telemetry — and you can do your own analysis on facility data rather than waiting for someone to hand you a chart. You treat measurement and verification as the core of the work, because in this business an unverified improvement is not an improvement.

Technically credible without needing to be the physicist. You can absorb enough of the underlying physics and modeling to make good scoping and sequencing decisions, ask the questions that expose risk, and explain the work accurately to a customer. You do not need to derive the models or write production code.

Bias to ship. Demonstrated ability to operate without bureaucracy in fast-moving environments. You ship; you do not stall. You are comfortable being the person who decides when nobody has told you the answer.

Appetite for scope. You want a role that grows faster than a title normally allows, and you are willing to take on responsibility ahead of formal structure to get it.

Nice to Have

• Experience with thermal engineering, CFD, HVAC systems, or power systems, whether formal or self-taught.

• Background in energy efficiency, retrofit, or optimization programs where results had to be measured and defended against a baseline.

• Experience standing up a delivery, professional services, or deployment engineering function at an early-stage company.

• Experience operating within regulated industries and familiarity with export control compliance and sensitive technology research and development.

• An engineering degree or equivalent hands-on technical background.

How We Work

We are engineering-led. Engineers own problems end to end, from spec to ship to on-call. We write contracts before logic, test against real systems instead of mocks, and favor simple designs that ship over clever ones that do not. Our development process is AI-native, and we expect the delivery side of the company to run on the same principles.

Location and Compensation

This role is based in Boston and is on-site at our Cambridge hub, with regular travel to customer facilities. We offer competitive compensation including salary, benefits, and meaningful early-stage equity. We evaluate delivery judgment, what you have shipped, and the leverage you create. Because of our export control obligations, some responsibilities may carry restrictions tied to citizenship or immigration status, which we will discuss transparently during the process. We are an equal opportunity employer and value diverse perspectives in building platforms for AI-driven discovery.