Principal Product Manager, Augmented Memory Grid (AMG)

Role overview Own the product direction and market adoption of a system that helps large language model inference workloads use memory and storage more efficiently. This technical product role connects customer needs with engineering decisions across GPU memory, networking, and distributed storage, with a focus on improving inference speed, capacity, and operating costs.

Responsibilities - Set the roadmap for cache offloading, memory tiering, and integrations with inference engines and serving platforms. - Work with engineering to turn performance constraints—such as response start time, throughput, and context length—into product requirements and architecture decisions. - Gather requirements from enterprise customers and infrastructure partners running production inference workloads; validate benchmarks and prioritize improvements. - Coordinate joint roadmap and certification work with accelerator and inference technology partners. - Define and track performance measures, including throughput and cache hit rates, and explain results to customers and internal teams. - Equip sales and field teams with technical positioning and support for enterprise opportunities.

Requirements - 10+ years of product management experience, including work in infrastructure, machine learning platforms, or developer-focused products. - Hands-on product or engineering experience with LLM inference serving and familiarity with caching, quantization, and batching approaches. - Understanding of production model serving, including context windows, multi-tenant workloads, and GPU scheduling. - Fluency in high-performance networking and distributed systems concepts such as RDMA, GPUDirect Storage, or NVMe-oF. - Experience working directly with large enterprise customers on requirements, production deployments, and service-level expectations. - Strong cross-functional leadership and communication skills, including the ability to explain technical trade-offs to varied audiences.

Nice to have - Experience at a GPU cloud, inference platform, or AI infrastructure company. - Familiarity with distributed file systems or object storage used in AI and machine learning pipelines. - Experience working with accelerator or silicon vendors.