M
Product Manager, AI Software Stack
Not specifiedNot specifiedProduct ManagementComputer Hardware
Define and deliver software products that help developers and cloud infrastructure teams compile and run AI models efficiently on a custom AI computing platform.
Responsibilities
- Translate customer, engineering, and business needs into product requirements, technical specifications, and feature definitions.
- Plan product milestones and releases, coordinating dependencies across engineering teams.
- Set requirements and roadmap for compiler pipelines and AI model toolchains.
- Lead integrations with machine learning frameworks and runtimes, including PyTorch, JAX, vLLM, SGLang, and Hugging Face Optimum.
- Define performance roadmaps for low-level operator libraries and optimized kernels.
- Set requirements for clustering and multi-node execution using tools such as SLURM, Ray, and LMCache.
- Advocate for developer experience and remove barriers to compiling and running software in customer environments.
Requirements
- At least 5 years of product management or software systems engineering experience, including shipped compiler technology, deep learning frameworks, or developer platform tools.
- Master’s or doctoral degree in computer science, electrical engineering, mathematics, or physics.
- Strong understanding of CUDA, Triton, high-performance C++ libraries, and parallel computing.
- Knowledge of inference-serving systems, batching, KV cache management, and latency-throughput tradeoffs.
- Familiarity with HPC scale-out infrastructure, distributed memory architectures, and high-performance networking.
- Experience in an accelerated software, cloud, or startup organization building complex AI compute systems.
Nice to have
- MBA or equivalent experience in technology commercialization or product strategy.
- Contributions to open-source AI infrastructure, compilers, or performance optimization libraries.
- Experience at a frontier AI lab or foundation model organization building large-scale distributed systems.