Databricks announced it has acquired Quotient AI, a startup focused on evaluation and reinforcement learning technologies for AI agents. Financial terms of the transaction were not disclosed.
The acquisition is aimed at improving how organizations deploy, monitor, and optimize AI agents in production environments. Databricks plans to integrate Quotient’s technology across its platform to strengthen its agentic AI capabilities and enable continuous evaluation and learning for enterprise AI systems.
Quotient AI was founded by engineers who previously worked on quality improvements for GitHub Copilot. The company built a platform designed to analyze full traces of AI agent behavior in production environments. By monitoring these traces, the system can detect issues such as hallucinations, reasoning failures, and incorrect tool usage.
The platform converts these signals into structured evaluation datasets and reward signals that can be used to improve agent performance through reinforcement learning. This process allows organizations not only to monitor agent behavior but also to systematically enhance performance over time.
Databricks said the technology will strengthen several core products on its platform.
- Genie, an AI agent that enables employees to interact with and analyze enterprise data through natural language.
- Genie Code, an autonomous AI agent designed to help data teams plan, build, and manage data engineering, machine learning, and analytics workflows.
- Agent Bricks, a system that allows organizations to build and scale high-quality AI agents using their own data.
As enterprises increasingly deploy AI agents into critical business workflows, the need for reliable evaluation and improvement systems has become more important. Databricks said combining Quotient’s technology with its AI research and platform capabilities will help organizations build agents that not only operate in production but also continuously improve based on real-world usage.
The Quotient AI team will join Databricks as the company continues developing tools for building, deploying, and optimizing AI-driven systems at scale.
