Post-Training Applied Researcher

Baseten is a company that specializes in powering AI inference for leading AI firms by combining applied AI research, infrastructure, and developer tools. The team focuses on enabling the deployment of advanced models into production, supporting clients with specialized AI needs. The company has recently secured significant funding and is expanding its platform to help engineers ship AI products efficiently.

The role of Post-Training Applied Researcher involves working closely with AI companies to fine-tune open-source models for specific tasks. Daily activities include designing training pipelines, creating evaluation environments, and translating real-world data into training signals. The person will also analyze training results, troubleshoot issues, and contribute to open-source libraries, with a focus on models used in domains like healthcare, legal, and coding.

Candidates should have practical experience with LLM fine-tuning and reinforcement learning, especially in designing reward functions and training environments. They need to be comfortable working across the entire ML pipeline from data collection to deployment, and have a good understanding of RL frameworks. The role offers a hybrid work setting in San Francisco and involves publishing research findings and collaborating with customers to improve models.