Technical Program Manager

Patronus AI, Inc.•San Francisco, CA

4d•$125,000 - $250,000•Onsite

About The Position

Patronus AI is a frontier lab developing simulation research and infrastructure to accelerate progress toward human-aligned AGI. We are on a mission to simulate all of the world’s intelligence. We are the team behind some of the earliest and most influential research in AI evaluation like FinanceBench, Lynx, SimpleSafetyTests, CopyrightCatcher, Humanity’s Last Exam, and more. We are formerly AI researchers and engineers from companies like Meta AI, Amazon AGI, and Google. Our customers include foundation model labs and Fortune 500 enterprises like Adobe. We are backed by top-tier investors like Lightspeed Venture Partners, Notable Capital, Stanford University, Noam Brown, Gokul Rajaram, and more.

Requirements

BS, MS, or equivalent experience in Computer Science, Engineering, or another technical / quantitative field, with 3+ years as a technical program manager, delivery lead, engineering program manager, or in a similar role on complex, multi-team technical programs.
A track record of shipping when requirements move after kickoff, without letting quality or the customer relationship slip.
Strong technical fluency, including comfort using AI tools, reading or reviewing code, and analyzing data or model outputs. You should be able to go deep enough into the work to have credibility with engineers.
Excellent organization and execution skills, with the ability to manage tasks, timelines, quality reviews, customer requirements, and cross-functional stakeholders. You are the person who always knows the current state.
Clear written and verbal communication skills, including the ability to translate customer asks into concrete plans with owners and dates.
Strong eye for quality and detail, with a bias toward catching edge cases, inconsistencies, and subtle failure modes before the customer does.
Proactive mindset, willingness to learn, unlimited energy, and relentless optimism.

Nice To Haves

Experience with reinforcement learning, agent evaluation, RL environments, or human-data / SME-sourced data pipelines.
Experience delivering to frontier labs or other research-driven customers with short turnaround expectations.
Experience standing up QA or review processes for software or data deliverables.

Responsibilities

Own delivery programs end-to-end: track work orders, deliverables, due dates, and owners across engineering, QA, and SME teams, and maintain the consolidated view the rest of the company plans against.
Manage scope changes mid-program. A previously committed app list doubles in size, or a customer redefines task difficulty on an active work order, and you update the plan, the pricing inputs, and our commitments without losing the thread.
Run the handoffs between GTM, SME sourcing, environment engineering, task generation, and QA. This is where work gets stranded today. Build the process that stops that: clear entry and exit criteria, and visibility into who is staffed on what.
Define what "ready to ship" means and hold the line on it. QA tickets are green before anything is marked complete, and tasks pass in the customer's harness, not just locally. We deliver work that already passes rather than delivering and triaging after.
Work directly with customers alongside account leads. Turn their asks into plans with clear outcomes, owners, and dates, and keep them current on progress, risks, and timeline confidence. Respond quickly; customers notice when we don't.
Get teams to time-bound their work and make estimates explicit ("expect this to take X hours, tell me if it takes longer"), then hold everyone, including yourself, to them.
Identify opportunities for automation and process improvements across the delivery pipeline and build them yourself.
Stay hands-on. Run agents in environments, review QA feedback and trajectories, and build small tools (trackers, scripts, Claude skills) that scale your own function.