Drimva

SambaNova

Principal PM, Product Strategy and Competitive

$216K–$260K

San Jose, California, United States · Principal · Full-time · Posted Sep 9, 2026

Experience
8+ years in specified fields.
Education
Bachelor's degree in engineering, computer science, economics, or another quantitative field, or equivalent experience.

About the job

Turns semiconductor and AI infrastructure market intelligence into product, pricing, positioning, and investment decisions, while building AI-powered research tooling.

What you'll do

  • Analyze research, competitor specifications, roadmaps, and benchmarks to quantify their impact on SambaNova’s costs, latency, and market positioning.
  • Build automated, human-reviewed pipelines to monitor sources, extract competitive intelligence, and draft recurring briefs.
  • Deliver weekly briefs, monthly deep dives, and quarterly thesis reviews to inform product and executive decisions.

What they ask for

  • 8+ years in semiconductor or AI infrastructure analysis, competitive intelligence, equity research, or corporate or product strategy at a chip, cloud, or AI infrastructure company.
  • Experience building quantitative market or economic models from primary inputs.
  • Fluency in accelerator and system architecture and its impact on inference performance and cost; track record translating competitor changes into decisions.
  • Hands-on with Python and LLM APIs; able to build or direct agentic research tooling.

Nice to have

  • Direct experience with SemiAnalysis models or comparable institutional research.
  • Experience in inference economics, rack- and cluster-level systems, capacity planning, or GPU economics.

Pay

$216,000 — $260,000 USD base salary

Tools

Python, LLM APIs, SemiAnalysis, InferenceMAX, ClusterMAX, NVLink, Ethernet, InfiniBand

AI focus

Build and continually improve agentic research pipelines to monitor sources, extract competitor specifications and claims, and automate first drafts of intelligence briefs.

Model-written summary · the original posting is authoritative

First seen Oct 10, 2026 · Last verified Nov 5, 2026 (UTC)