Baseten posts post‑training role
Baseten has an open Post‑Training Research Engineer role, which paired with the Sameer hire points to ramped investment in post‑training and deployment workflows. That hiring pattern suggests they’ll be increasing experiment and production GPU needs shortly. (linkedin.com)
The Post‑Training role’s listing calls for PhD‑level research depth, first‑author papers at top venues, and building experiments at “multi‑node, 1T+ parameter” scale — responsibilities tied explicitly to production experimentation rather than purely theoretical work. (jobgether.com) Baseten’s public jobs board shows parallel hiring across model performance and infrastructure — openings titled “Software Engineer — GPU Kernels”, “GPU Networking & Distributed Systems”, and “Model Performance” with posted salary bands up to ~$360K. (jobs.ashbyhq.com) Baseten announced Sameer Paranjpye as a senior engineering hire; the company says he spent nine years as VP of Engineering at Databricks and previously founded a startup, bringing scaling experience for developer‑facing platforms. (baseten.co) The company closed a $300M Series E at a $5B valuation on Jan. 23, 2026, with NVIDIA named as an anchor investor in the round. (businesswire.com) In December 2025 Baseten acquired Parsed, a reinforcement‑learning team focused on post‑training and continual learning for large models, and the Parsed engineers were already working with Baseten customers prior to the buy. (businesswire.com) Baseten’s Series E announcement states the business saw inference volume grow 100x in the prior year and lists customers such as Cursor, Notion, and Abridge as examples of high‑volume deployments. (baseten.co) Multiple role descriptions require publishing at NeurIPS/ICML/ICLR and “moving from theory through implementation to empirical results,” indicating an expected pipeline from post‑training research to deployable, production‑grade experiments. (jobgether.com)
