About the job
The Data Agent team within AI Research is dedicated to pioneering frontier enterprise data agents. Our research focuses on developing agents capable of autonomous planning, code generation, and multi-step workflow execution within intricate enterprise settings. The team's primary technical pillars include post-training enhancements, harness design, agentic reinforcement learning (RL), and the construction of specialized RL environments. As a member of this team, you will bridge the gap between cutting-edge AI research and product development by shipping direct improvements to Genie, Databricks' agent product that provides significant value to our user base.
Responsibilities
Develop the best post-training recipes to train enterprise Data agents, that are capable of autonomous planning, code generation, and multi-step workflow execution within intricate enterprise settings.
Partner closely with product teams to turn prototypes and research ideas into the best agentic experience for Databricks users.
Build systems that help the agent discover and use relevant lakehouse context, including tables, notebooks, code, and cell outputs, to produce more accurate and useful results.
Raise the technical bar for the team through strong design, execution, debugging, and mentorship, helping shape the long-term direction of agentic experiences at Databricks.
Qualifications
Minimum
BS, MS, or PhD in Computer Science or a related field.
2+ years of experience in an applied research environment, with a track record of shipping research prototypes to production.
Experience in LLMs, agents, reinforcement learning, post-training workflows.
Preferred
No preferred qualifications listed.