Prior Labs | Berlin, Freiburg, NYC | ONSITE | Full-time | https://priorlabs.ai
Core
Building foundation models for tables, specifically TabPFN, which predicts on whole datasets in one forward pass without training or tuning.
Role type
Infrastructure Engineer, ML Engineer (Backend), Full Stack Developer
Builds
Multi-cluster GPU training infrastructure, inference API, and the TabPFN product platform
Domain
AI/ML, Foundation Models, Table Data
Deliverable
production ML models
Required skills
GPU cluster management, Slurm, GCP, Python, TypeScript, React, FastAPI, API design, latency optimization, memory management
Preferred skills
Experience with open-weight models, multi-provider cloud infrastructure
Technologies
Slurm, GCP, Python, TypeScript, React, FastAPI
Responsibilities
Own multi-cluster GPU training infrastructure and compute budget; manage end-to-end inference latency, batching, and memory for million-row inputs; build product features connecting data scientists, agents, and models
Seniority
Mid-level to Senior IC
