Description
At NVIDIA, we meet customers where they are on their AI journey on our GPUs, building best-in-class frameworks in open-source software and supporting a robust ecosystem of other open-source software frameworks.
As a Senior Product Manager for AI Platform post-training, you will be responsible for building products that enable post-training and reinforcement learning (RL) at scale on NVIDIA GPU and CPU systems.
Your goal will be to enable researchers and operators to be successful on the NVIDIA Platform and push the boundaries of what is possible in their research and research-to-production.
Post-training is a de facto workflow amongst our customers and partners, and post-training and RL are no longer single-system workloads but require distributed systems to work reliably at scale.
As a Product Manager, you will:
- Create and optimize pre-training/inference and post-training frameworks for researchers and production model builders
- Develop product strategy, roadmaps, and go-to-market plans
- Collaborate with internal and external customers to build product-based roadmaps for the end-to-end machine learning lifecycle
- Work with leadership to align with and drive company strategy
We are looking for someone with:
- Experience in designing and scaling training/inference post-training and optimization software (e.g., Torchtitan, FSDP)
- Demonstrable knowledge of generative AI or machine learning concepts, particularly around model training, performance optimization, inference, and software development and delivery
- Experience with large-scale distributed systems
- A BS or MS degree in Computer Science, Computer Engineering, or similar experience (or equivalent experience)
- 10+ years of technical product management experience at a technology company
- Strong communication and interpersonal skills
To stand out from the crowd, you should have:
- Background with leading open-source RL frameworks like VeRL
- Experience with orchestration and reliability at scale
- Experience working on open-source and GitHub-first developer products with deep customer interactions
- Knowledge of GPU architecture, hardware/software co-design, and performance profiling
You will also be eligible for equity and benefits.