Technical Program Manager, Model Alignment and Deployment

About the role

Model Alignment and Deployment is a critical, cross‑functional effort that transforms pretrained language models into safe, engaging, and scalable products. As a Technical Program Manager you will be the operational backbone that connects research, safety, user experience, and production teams to deliver high‑impact AI capabilities.

Key responsibilities

• Lead planning and execution of cross‑functional programs spanning data collection, annotation pipelines, alignment workflows (RLHF, DPO, Constitutional AI), safety guardrails and model serving.
• Serve as connective tissue between Post‑Training, Safety Engineering, Trust & Safety, ML Infra, UXR and Product, translating priorities into executable roadmaps.
• Develop and maintain custom evaluation frameworks to track model performance, user satisfaction and safety metrics.
• Drive visibility into data pipeline health, annotation quality, training run progress and deployment readiness, identifying bottlenecks and improving tooling.
• Partner with research, safety, product and UXR leadership on prioritization, sequencing and trade‑offs.
• Build and refine operational patterns, ontologies and frameworks for scaling new capabilities, from prompt engineering to safety guidelines.
• Own external partner relationships with annotation vendors, evaluation tooling providers and data partners.

Required profile

• 5+ years of experience in technical program management, research operations or product execution in fast‑moving AI/ML environments.
• Deep familiarity with post‑training and alignment concepts such as supervised fine‑tuning, RLHF, AI safety frameworks and LLM evaluation.
• Proven ability to lead complex, multi‑team programs in ambiguous, rapidly evolving settings.
• Strong analytical mindset with experience using data and user insights to measure program health.
• BS in a quantitative, scientific or technical field (MS/PhD a plus).

Required skills

• SQL
• Python
• Kubernetes
• Docker
• Cloud computing platforms
• Distributed training and experiment tracking
• ML infrastructure and model serving systems
• Data pipelines and annotation platforms
• ML evaluation tooling and human‑in‑the‑loop workflows