About The Position

At Microsoft AI, we build the frontier models and products that reach millions of people, from pretraining through post-training, evaluation, and launch. Our Technical Program Managers are a centralized team, embedded deeply with the research, product, and engineering stakeholders they support, driving execution across every pillar and workstream. This role sits inside that pipeline, helping turn ambitious research into shipped, trusted systems. We are looking for a Technical Program Manager with deep, hands-on technical program management experience to support model development at MAI. You'll partner closely with research, product, and applied science leads, bringing enough technical fluency in model training, evaluation, and infrastructure to engage credibly with researchers and engineers, and enough program rigor to keep complex, fast-moving work on track. As TPM, you will own end-to-end execution across the model development lifecycle, partnering with research and applied science leads to shape goals, priorities, and success metrics, and managing planning, milestones, dependencies, and risk across training, evaluation, and launch. You'll be the connective tissue between research, engineering, product, and safety teams, translating fast-changing research priorities into clear plans, reasoning about trade-offs in data, compute, and evaluation methodology, and surfacing risks to leadership before they become blockers.

Requirements

  • Significant experience in technical program management, with a track record of delivering complex, cross-functional programs, ideally spanning pretraining, post-training, or evaluation.
  • Technical fluency to reason about model training, evaluation methodology, data, and infrastructure trade-offs, and engage credibly with researchers and engineers.
  • Strong written and verbal communication skills, with a track record of driving alignment and influencing without formal authority.

Responsibilities

  • Own end-to-end execution of complex technical programs across the model development lifecycle, partnering with research and applied science leads to translate research goals into clear plans, milestones, and success metrics.
  • Reason about technical trade-offs in model training, evaluation, data, and infrastructure, and coordinate across research, engineering, and product to keep training and evaluation work unblocked.
  • Communicate program status, risk, and trade-offs clearly to stakeholders and leadership, bringing lightweight structure to fast-moving, ambiguous research work.
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service