GTM Research Intern (US) US GTM San Francisco

Position: GTM Research Internship (US) Location: San Francisco, CA (In-person) Type: Summer 2026 intern (10 to 12 weeks, full-time) About VESSL AI AI researchers spend more time fighting for compute than doing research. Quota walls, rigid contracts, and fragmented supply across clouds and clusters mean teams are stuck waiting on a single provider that can't meet their needs. VESSL AI was built to fix that. We provision and route GPU capacity across multiple clusters and providers, giving AI teams one platform to access, scale, and manage compute, from a single GPU to thousands. Multi-provider procurement so you're never locked to one source. Full-stack cluster operations from build to production. Deep domain expertise in large-scale training optimization. We serve top research labs, academic institutions, AI startups, and enterprise teams globally. Customers include an Artificial Analysis Top 15 foundation model team, Physical AI startups, and research groups at Stanford, UC Berkeley, UMN, UCLA, NYU, and more. Founded in Seoul, South Korea, now expanding to San Francisco. Series A, ~40 people, scaling from 3,000 to 10,000 GPUs by Q4 2026. About the Role The GPU supply constraint has scrambled the AI infrastructure market. Demand is fragmented across thousands of teams running serious compute: frontier labs, AI startups at every stage, Physical AI companies, academic labs pushing on capacity, AI for science teams, enterprise ML platforms. You will be the person who maps that demand. This is greenfield research work at the highest-leverage point in our GTM motion. Without a clean map of who actually needs compute, every other GTM function is guessing. You will replace the guessing with data. You will work directly with the US GTM Lead. Your demand intelligence will feed the entire US pipeline. Who We Expect Currently pursuing a Bachelor's at a top program, or recently graduated Strong research instincts: finds signal where others see noise and validates hypotheses fast Deep market curiosity: actively tracks where the AI/ML infrastructure landscape is headed and how it's evolving AI-native: already uses n8n, Claude Code, or equivalent daily Operational ownership: ships the work, debugs the data, does not wait to be asked On-site SF for the full 10 to 12 weeks, Summer 2026 #J-18808-Ljbffr VESSL AI, Inc.