Databricks has launched Agent Bricks, an automated solution designed to facilitate the creation of AI agents tailored to specific business needs.

This tool allows users to input a “high-level” description of the desired task and connect their enterprise data, with Agent Bricks managing the subsequent processes.

The service, now available in Beta, is optimised for various industry applications, including structured information extraction, knowledge assistance, text transformation, and multi-agent systems, the company said.

Agent Bricks employs advanced research methodologies from Mosaic AI Research to generate domain-specific synthetic data and task-aware benchmarks.

This approach enables automatic optimisation for both cost and quality, streamlining the development process and enhancing production-level accuracy.

The integration of governance and enterprise controls allows teams to transition from concept to production efficiently, eliminating the need for disparate tools.

The functionality of Agent Bricks includes automatic generation of task-specific evaluations and LLM judges, the creation of synthetic data that mirrors customer data, and a comprehensive search for optimisation techniques.

Users can select the iteration that best balances quality and cost, resulting in a production-ready AI agent capable of delivering consistent output, the company’s statement added.

Agent Bricks supports various customer use cases across multiple sectors. For instance, the Information Extraction Agent converts documents into structured data, while the Knowledge Assistant Agent provides accurate answers based on enterprise data.

The Multi-Agent Supervisor facilitates the integration of multiple agents for complex tasks, and the Custom LLM Agent allows for tailored text transformations.

Databricks CEO and co-founder Ali Ghodsi said: “For the first time, businesses can go from idea to production-grade AI on their own data with speed and confidence, with control over quality and cost tradeoffs.

“No manual tuning, no guesswork and all the security and governance Databricks has to offer. It’s the breakthrough that finally makes enterprise AI agents both practical and powerful.”

In addition to Agent Bricks, Databricks has introduced several features at the Data + AI Summit, including support for serverless GPUs, enabling teams to fine-tune models and run workloads without managing GPU infrastructure.

The release of MLflow 3.0, a platform for managing the AI lifecycle, allows users to monitor and optimise AI agents across various environments.

In May 2025, Databricks announced the acquisition of Neon, a serverless Postgres database company.