Databricks debuts Genie Code & snaps up Quotient AI

Databricks has launched Genie Code, an autonomous coding agent for data engineering, data science, and analytics teams, and has acquired agent evaluation specialist Quotient AI.

Genie Code is positioned as a shift from AI tools that offer snippets and suggestions to systems that carry out multi-step work across production data environments. Databricks says it can build data pipelines, debug failures, ship dashboards, and maintain production systems.

Data agent focus

Genie Code sits within Databricks' Genie product line, which is designed for conversational access to organisational data. Genie enables knowledge workers to "chat with their data" using context and semantics held in Unity Catalog. Genie Code extends that approach to technical teams, turning an initial request into deployed data work.

A central part of the pitch is access to enterprise context. Many coding agents struggle with data tasks because they lack information such as lineage, usage patterns, and business semantics. Databricks says Genie Code uses Unity Catalog to operate within governance policies and access controls.

Databricks says the system "reasons through problems" by producing multi-step plans, writing code, and validating outputs. It also maintains deployed artefacts, while leaving key decisions with human users.

Product scope

Databricks outlined several expected use cases, including machine learning development. It says the agent can handle "full ML workflows end-to-end," log experiments to MLflow, and fine-tune serving endpoints.

Another use case is data engineering across staging and production. Databricks says Genie Code designs workflows that account for differences between environments, supports change data capture, and applies data quality expectations.

Databricks also highlighted ongoing operations tasks. It says the agent monitors Lakeflow pipelines and AI models, triages failures, and investigates anomalies. The system can also analyse "agent traces" to address hallucinations and tune resource allocation.

Databricks says the agent improves with use through persistent memory, updating internal instructions based on past interactions and coding preferences.

In its own testing on "real-world data science tasks," Databricks reports Genie Code increased the success rate of leading coding agents from 32.1% to 77.1%.

Ali Ghodsi, co-founder and CEO of Databricks, framed the release as an extension of recent changes in software development tooling. "Software development has shifted from code-assistance to full agentic engineering in the past six months," Ghodsi said. "Genie Code brings this revolution to data teams. We're moving from a world where data professionals are assisted by AI to one where AI agents do the work, guided by humans. We are calling this Agentic Data Work. It will fundamentally change how enterprises make decisions."

Customer references

Databricks cited early usage at SiriusXM, saying Genie Code is being used for notebook authoring, SQL development, and pipeline debugging.

"At SiriusXM, Genie Code supports everything from authoring notebooks and complex SQL to reasoning through table relationships and debugging pipelines," said Bernie Graham, VP of Data Engineering at SiriusXM. "It acts as a hands-on development partner that helps our data teams deliver high-quality work in less time."

Repsol also shared comments on its use of the product, highlighting workflow handoffs to an AI system that reflects internal context and libraries.

"Genie Code changes how our data teams operate," said Emilio Martín Gallardo, principal data scientist, Data Management & Analytics at Repsol. "Instead of stitching together notebooks, pipelines, and models manually, we can hand off complex workflows to an AI partner that understands our data, governance, business context, and internal libraries such as Repsol Artificial Intelligence Products. It accelerates everything from time series forecasting to production deployment, without sacrificing rigor or control."

Quotient acquisition

Alongside the product launch, Databricks said it acquired Quotient AI, which focuses on evaluation and reinforcement learning for AI agents. Databricks plans to embed continuous evaluation into Genie and Genie Code.

According to Databricks, Quotient's technology monitors agent performance, measures answer quality, flags regressions, and identifies failures. Databricks says these signals feed a reinforcement learning loop that improves agents over time.

Databricks also noted that Quotient's founders previously worked on quality improvements for GitHub Copilot, and said it expects the acquisition to strengthen its ability to keep agent behaviour under review once deployed in production environments.