Agent Bricks: The Governed Enterprise Agent Platform

The basic agent pattern is familiar by now: a model connected to tools, reasoning, and taking actions. But building the loop is not the hard part. The hard part is making enterprise agents work on real business data, under real permissions, with real consequences.

The most valuable agents are defined by how deeply they connect to your business: customer records, operational systems, internal policies, and institutional knowledge. A financial services agent reviewing loan applications and applying company underwriting policies is valuable because it operates in a business context, not because of the model or framework alone. That context is what makes agents useful, and what makes them difficult to run in production. Agents need to understand what data means, operate under the right identity and permissions, and work across models without locking teams into a single vendor.

This is where most teams get stuck. Most agent products give you pieces, not a platform.

This is why we built Agent Bricks. Agent Bricks is Databricks’ enterprise agent platform for building, deploying, and governing agents that operate on your business data, end to end. It unifies model access, execution, governance, and context so teams can run agents reliably in production.

Thousands of organizations across financial services, retail, healthcare, and technology have deployed production agents at scale on Agent Bricks, including Workday, Virgin Atlantic, Zapier, EchoStar, and AstraZeneca. Teams are building agents that deliver continuous market analysis to hundreds of analysts, orchestrate workflows across supply chain, procurement, and R&D systems, solve employee requests for complex service tasks automatically, and detect and resolve anomalies in marketing campaigns before advertising dollars are wasted.

Today, we are announcing the general availability of Document Intelligence and Custom Agents, along with new capabilities across the platform, including AI Gateway, to help you build, govern, and secure enterprise agents grounded in rich context from your data.

"With Agent Bricks, we’re not building one-off AI projects, we’re building an enterprise AI fabric. Interoperability, identity-first security and governance were designed from day one, so our agents behave like any other mission-critical system, not a science experiment." — William Acosta, Head of Agentic AI Engineering, EchoStar

The Agent Bricks platform

Running agents in production requires more than a model and tools. It requires a platform. Three things define the Agent Bricks platform:

Open and multi-AI. To build useful agents and agentic applications, teams need to work across multiple model providers and frameworks to choose the right models, use the right tools, and manage access, cost, and reliability. Agent Bricks natively supports frontier models and popular coding agents like Cursor, Codex, and Claude Code through a single API, with built-in routing, fallback, and cost optimization. It also supports building and deploying agents with major frameworks such as LangGraph and OpenAI Agents SDK. This allows teams to switch models or integrate external agents without rebuilding their systems. Today, 63% of customers route tasks across two or more model families, ensuring agents remain flexible and resilient as models evolve. 

Unified governance. Most platforms govern the agent, which tools it can call, and what permissions it holds. Agent Bricks governs the agent and everything it interacts with in a single system. With Unity Catalog and AI Gateway, access to data, models, and external MCPs is managed and observed in one place, with identity enforced end to end. Agents inherit user identity through on-behalf-of token passing, so they can only access what the user is authorized to use. Whether querying your lakehouse or calling external APIs, the same permissions, auditing, and routing apply across every interaction. This ensures every agent interaction is secure, observable, and consistent.

"Agent Bricks gives us a structured way to coordinate multiple data intelligence endpoints in a single system. Instead of hard-coding routing logic, we can guide how the agent prioritizes Genie and governed data in Unity Catalog through clear instructions. That makes it much easier to build an internal ‘ask data’ experience that’s flexible and reliable as it evolves.” — Alvaro Martin, Sr. Data Engineer, Zapier

Accurate because it understands business context. Agent accuracy depends on more than model quality. Agent Bricks uses Unity Catalog metadata, including schema, business definitions, lineage, permissions, and data quality signals, to improve how agents reason and act. This context is embedded directly into retrieval and planning, delivering 70% higher accuracy than standard RAG and a 30% improvement in multi-step workflows. For structured data, Genie Spaces leverage the semantic layer so agents reason over business definitions, not raw column names. This means agents return answers aligned to how your business actually operates, not just what the data says. This is what turns models into systems that understand your business.

What's new

Today’s releases expand what teams can build on Agent Bricks across multi-AI, governance, and enterprise context:

Multi-AI and agent orchestration

• Custom Agents on Apps (GA). Build and deploy agent applications with any model or framework, with full lifecycle support and serverless compute. Native integration with Lakebase provides memory, conversation history, and state for long-running workflows.
• Supervisor Agent (GA). Orchestrate multiple agents and tools into a single workflow. Define the task and connect your systems. The supervisor coordinates execution across models and tools.
• Web Search in Foundation Model API. Ground agent responses with real-time information from the web using native provider search capabilities.

Governed access across tools, models, and data

• AI Gateway. A unified layer to manage and govern access to models, coding agents, and now MCP-connected tools. It enforces identity, permissions, and observability across every interaction, so agents operate securely across your models, tools, and APIs. Now also with guardrails to detect and mitigate risks like PII exposure, unsafe content, prompt injection, data exfiltration, and hallucinations, with customizable options for various security needs.
• Managed OAuth MCP Connectors. Securely connect external services like GitHub, Atlassian, and Glean as governed tools. Credentials are managed centrally, so agents can access systems without exposing secrets.

Enterprise context that makes agents accurate

• Document Intelligence (GA). Extract and structure data from unstructured documents like contracts, invoices, and reports, turning PDFs into queryable knowledge without custom pipelines.
• Knowledge Assistant (GA). Automatically ingest enterprise documents and make them accessible to any agent, with retrieval that incorporates system context, metadata, and user constraints.
• Agent Mode in Genie Spaces. Move from single-turn Q&A to multi-step reasoning and analysis over your data, enabling agents to plan, explore, and answer complex business questions.
• CLEARS Framework for Agent Quality with MLflow. Evaluate agents across correctness, latency, execution, adherence, relevance, and safety with a standardized framework in MLflow for production quality.

Build on Agent Bricks today

The challenge is no longer building the agent loop. It is building everything around it: identity that cannot be bypassed, credentials that do not leak, model routing that avoids lock-in, business context that makes results correct, and observability that shows what every agent did and why.

Most platforms focus on pieces of this system. Agent Bricks brings it together.

It is a multi-AI, governed platform for agents operating on enterprise data, designed to run reliably in production from day one. And we are continuing to expand that system through a growing ecosystem of integration partners, including Accenture, Atlan, Arize, Capgemini, Celebal, Collibra, Daitaiku, Deloitte, EY, Glean, Infosys, LlamaIndex, Lovelytics, LTIMindtree, Monte Carlo, Omni Analytics, Qlik, Retool, Sigma Computing, Slalom, Tiger Analytics, Tredence, and Wipro, extending Agent Bricks across the tools and services enterprises rely on every day.

• Get started with the documentation →
• Start a Databricks trial →