Agent Bricks
Recent items mentioning Agent Bricks across the Databricks ecosystem — releases, news, videos, and community Q&A. Updated hourly.
Databricks is pushing Agent Bricks as the governed on-ramp for agent development: it now natively supports Model Context Protocol and Unity Catalog for grounding tool-calling agents in enterprise data 3, and pairs with Databricks Apps and Omnigent to build, redeploy, and govern custom agents with MLflow trace tracking and Unity AI Gateway security policies 24. It's also becoming a standard downstream consumer of platform data features, with the newly GA Variant type now integrated for feeding semi-structured data into Agent Bricks workflows 5.
Written August 2026 from the 5 most recent items mentioning Agent Bricks at that time. It refreshes when this topic next has enough new material. Click any [N] to jump to the source.
How energy teams turn theft detection into governed action with Genie and AI business processes
A Databricks App orchestrates the full lifecycle of energy theft detection—turning ML-flagged suspicious accounts into prioritized investigations, dispatch-ready reports, and tracked recovery workflows, with Lakebase keeping live case state and recovery totals. Genie One, Unity Catalog, Unity Gateway, and Agent Bricks tie this together on one platform, delivering trusted metrics, governed AI usage, and automated executive reporting.
Learn Databricks Agent Bricks | Build Enterprise RAG Agents
NewsBuilding Agents on Databricks with Custom Apps and Omnigent
This video demonstrates how to build, update, and govern custom AI agents on Databricks using Agent Bricks, Databricks Apps, and Omnigent. The tutorial shows how to integrate Model Context Protocol servers, track execution with MLflow traces, schedule automated agent tasks, and manage security policies through Unity AI Gateway.
What is Tool Calling?
Tool calling transforms basic chatbots into action-oriented AI agents by executing a structured loop to interact with external tools, APIs, and systems. Databricks Agent Bricks provides a governed environment to build these agents grounded in enterprise data, featuring native support for the Model Context Protocol and Unity Catalog governance.
TutorialsBuilding Agents on Databricks with Custom Apps and Omnigent
The video demonstrates how to build, update, and govern a store operations AI agent on Databricks using Model Context Protocol servers and custom apps. It shows how to use Omnigent and CodeX to add new context and tools, redeploy the application, and manage governance and traces through the Unity AI gateway.
Ingest semi-structured data faster and more efficiently with Variant - Now Generally Available
Databricks Variant is now Generally Available, enabling teams to achieve up to 30x faster reads on semi-structured data while handling unpredictable schema changes without pipeline updates. The feature is broadly integrated across the platform, supporting data workloads like Auto Loader and Spark Declarative Pipelines alongside AI tools like Agent Bricks and AI Functions.
AI-Enabled Advisory Services for Higher Education
Databricks has introduced a GenAI-powered solution that automates and scales the quality review of higher education advisory calls by transcribing conversations, scoring performance against institutional rubrics, and surfacing insights. The blog post demonstrates this architecture on a single, governed platform, providing code notebooks and a Genie space for natural language data exploration.
NewsWhat’s coming next to Free Edition
Databricks announces the availability of Genie, GPUs, Agent Hooks, Lakehouse, and Lake Flow Designer on its Free Edition. This update provides virtually all of Databricks' production platform features for free, enabling users to learn and build data and AI projects.
Agentbricks Knowledge Assistant - Facing Issue with Agent Creation and Source Visibility
EventsWhat it takes to scale agents in the enterprise: context, control and choice
Databricks CEO Ali Ghodsi outlines enterprise AI scaling strategies centered on organizational context, open choice, security control, and cost management. The session demonstrates developer tools like Agent Bricks, the Omnigent open source project, and Genie code alongside enterprise partnerships.
What if the answer was already in your data?
Kythera Labs' AI agents, built on Databricks, now provide health system leaders with governed, trustworthy answers to strategic questions from 339 billion claims. A Louisiana health system saw 150% more visibility into patient encounters and $3.8M in estimated annualized value in 10 days.
EventsRecap of product announcements from Data + AI Summit 2026 | Day 1
Databricks announced several new products and features at the Data + AI Summit 2026, Day 1, including the Genetic Data Foundation, Lakehouse RT, Lake Base with disaster recovery, Lake Flow, Genie Ontology, Unity AI Gateway, Omnigent, and various Genie agents (Genie 1, Genie Code, Genie Agents). They also introduced new applications like Lake Watch for SIM and Customer Lake for CP.
What’s coming next to Free Edition
Databricks Free Edition now includes every core practitioner feature, expanding with Genie Code, GPUs, Lakebase, Lakeflow Designer, and Agent Bricks. This gives users a complete, free toolkit for building end-to-end data and AI projects.
Agent Bricks: Data + AI Summit 2026
Last year at the Data + AI Summit, we launched Agent Bricks, ushering in a new way...
Unlocking semantics for AI: How Mercedes-Benz Korea built trusted “Talk to Data” at scale
Mercedes-Benz Korea built a trusted "Talk to Data" solution at scale by making 500+ KPI definitions available in an AI-ready semantic layer on Unity Catalog metric views, accelerating the transition with an automated DAX-to-Metric-View transpiler. This governed semantic layer supports both existing BI and new "Talk to Data" experiences, with Genie and Agent Bricks providing consistent answers and shaping a playbook for persona-based AI agents across markets.
Self Learning Testing Framework using Experiment Tab in Agent Bricks
3x Faster Search: Parallel Test-Time Scaling with Instructed-Retriever-1
Instructed-Retriever-1 now delivers 3x faster search for Agent Bricks Knowledge Assistant. This parallel test-time scaling update also improves quality for Databricks practitioners.
Bring Databricks into Kiro IDE with the AI Dev Kit Power
The Databricks AI Dev Kit Power now offers a one-click setup to integrate Kiro IDE with the full Databricks platform, providing AI-assisted development grounded in your workspace's Unity Catalog metadata. This new path, alongside a lighter PAT-based option, ensures your AI assistant writes SQL with actual columns and respects all row, column, and tag-based grants.
NewsBanks' Secret Weapon Against Money Laundering: Multi-Agent AI
Databricks demonstrates a multi-agent AI solution for Anti-Money Laundering (AML) operations, significantly reducing false positives and accelerating investigation cycles from hours to minutes. The platform unifies siloed systems, employs specialized AI agents for analysis and recommendations, and offers AI-assisted SAR generation and executive-level reporting with natural language chat.
Multi-Agent Supervisor for Hybrid Retrieval with Agent Bricks and MLflow
The provider adds service principal Git credential management via principal_id on databricks_git_credential and enables permission management for Agent Bricks resources. Key fixes include metastore external_access_enabled now properly sent in PATCH requests, vector search index timeout increased to 75 minutes and made configurable, and workspace_id now accepting connection IDs alongside numeric workspace IDs.
Announcing Lakebase Change Data Feed (CDF)
Lakebase Change Data Feed (CDF) is now in Public Preview, eliminating pipeline sprawl from operational databases by exposing every table's changes through Unity Catalog Managed Tables. This enables native CDC governed end-to-end without sidecar infrastructure, allowing operational data to function as the native Bronze layer in the medallion architecture.
Document Intelligence on Databricks
80% of enterprise data is locked inside PDFs, scans, emails and contracts and most teams still treat it as someone else's problem. Document Intelligence on Databricks changes that. One SQL function (ai\_parse\_document), governed by Unity Catalog, integrated with Lakeflow for ingestion, Agent Bricks for structured extraction, and Vector Search for RAG. No stitched-together OCR vendors, no brittle Python glue, no separate platform to govern. I put together with [Archika Dogra](https://www.linkedin.com/in/archikadogra/) a walkthrough showing how it actually works end-to-end from a folder of raw PDFs to queryable Delta tables and downstream agents. ▶️ [https://youtu.be/sdG73gI143c](https://youtu.be/sdG73gI143c) Curious to hear what use cases you're tackling invoices, contracts, claims, technical docs? Drop them in the comments.
TutorialsBuilding Enterprise-Ready Agents using Agent Bricks
Databricks Agent Bricks is a unified platform designed to help enterprises build and manage AI agents, addressing challenges like low-quality reasoning on proprietary data, lack of governance, and fragmented toolchains. It demonstrates how to create knowledge assistants for unstructured data and AI Genies for structured data, integrating with Unity Catalog for governance and MLflow for observability and evaluation.
From "What Happened?" to "What Will Happen?"
Conversational BI now delivers predictive answers in seconds, not days, by fusing Genie for dynamic feature engineering with TabPFN for zero-training prediction, orchestrated by Agent Bricks. This self-assembling pipeline eliminates data science bottlenecks for business users, providing a governed experience backed by Unity Catalog and MLflow.
Agent Bricks in Action: Automating Insurance Underwriting with a Supervisor Agent-Led Architecture
Genie code or AI Dev kit?
Hey everyone — I’m planning to build a quick demo using Agent Bricks with a lightweight frontend through Databricks Apps. For development, would you recommend using Genie Code or AI Dev Kit? Curious which one people have found better for rapidly building and iterating on Databricks-native AI applications.
TutorialsHow to Build an AI Security Governance Hub with Agent Bricks
Databricks Agent Bricks enables building an AI Security Governance Hub by transforming static security playbooks into adaptive multi-agent systems. The video demonstrates combining a knowledge assistant for unstructured documents and a Genie space for structured data into a supervisor agent, then details how to tune and monitor these agents for improved performance and data privacy.
Agent Bricks | A Pilot to Production Series - Manufacturing
Agent Bricks | A Pilot to Production Series - Healthcare & Life Sciences
NewsDatabricks in 3 minutes. The unified data and AI platform, explained.
Databricks unifies diverse data sources into a single data lake, providing a governed platform for analytics and AI. It offers capabilities like fine-grained access control, natural language querying with AI, and company-wide intelligent agents.
Show HN: Recursant – service mesh for governing AI agents
Hello, I have just released Recursant to the public. I have been working on it for a while. It is a control plane for governing AI agents across stacks. It provides full observability, guardrails, and control on the network level by routing all traffic through a side car. Problem statement: many large, regulated enterprises (think banks, telcos) have one engineering team on LangGraph, another on CrewAI, marketing on AgentForce, and data teams on Databricks Agent Bricks. They need their agents to talk to each other with consistent policy enforcement, one audit trail, and a single set of guardrails, yet allowing different functions to run on their own stacks. Recursant solves that problem using the service mesh pattern Recursant has two components: a registry and the mesh . The registry contains all live agents. The mesh uses sidecars to route traffic and enforce on the network layer. Aim is for Recursant to provide a real-time EU AI Act Annex IV compliance, so it is not generated from static documents. This saves time and effort for large enterprises subject to the requirement. Linmitations: - Recursant currently plugs in to CrewAI, Langgraph, and n8n . The aim is to support proprietary platforms such as ServiceNow and AgentForce as much as psosible. - The Recursant SDK still needs work to support as many agents as possible - I would also like to provide support for some of the 'personal agent' platforms such as OpenClaw, NanoClaw, and Hermes - Only tested on k8s, not public cloud - Documentation is sparse and needs to be developed. I hope this project is useful to some of you. --- top comments --- [goodra7174] clawdlinux.org building the Kubernetes-native runtime that AI agents call to provision their own secure execution environments — the missing infrastructure layer for every enterprise that can’t send agent data to the cloud #Ycombinator #buildinpublic #startups
[PARTNER BLOG] Revolutionizing Enterprise Data with Databricks Agent Bricks: Tale of Two Industries
Agent Bricks | A Pilot to Production Series - Retail
How to transform document activation workflows with Genie and Agent Bricks
Databricks shipped a solution combining AI/BI Genie, Agent Bricks, and Unity Catalog to automate document activation workflows. This enables multi-agent orchestration for extracting, processing, and activating data from diverse documents, improving efficiency and governance.
NewsDatabricks Apps vs Model Serving: Authentication, Cost, and Performance Compared
Databricks Apps are now the recommended first choice for deploying agents due to their flexibility in handling full-stack applications with multiple components, offering faster iteration and local testing compared to Model Serving. Model Serving remains suitable for use cases prioritizing high QPS, governance features like AI Gateway, inference tables, and guardrails, or when scaling to zero is acceptable for cost optimization.
NewsTurbo-Charge your Agents with instant MCP in Databricks
The video demonstrates how to use Model Context Protocol (MCP) in Databricks to give AI agents "superpowers" by enabling them to interact with various tools and data sources. It shows how to easily set up MCP servers within Databricks to connect agents to Unity Catalog functions, vector search, external APIs, and even marketplace MCP services, all without extensive coding.
EventsDAIS25 Keynote Day 2 Sizzle
Databricks announced a free edition of its platform, allowing users to access a slice of Databricks forever without a credit card. The company also showcased Agent Bricks for building production-ready AI agents and Databricks Apps for secure data intelligence applications.
TutorialsBuild a Databricks Knowledge Assistant in less than 10 steps! with Agent Bricks | Advancing AI
The video demonstrates building a Databricks knowledge assistant using Agent Bricks, a no-code platform for creating agentic systems. It shows how to upload company policy PDFs to Unity Catalog, configure a knowledge assistant to answer questions based on these documents, and deploy it as a Streamlit chatbot application.
EventsCo-ordinating AI/BI Genie Spaces with Databricks Agent Bricks
Agent Bricks is a low-code tool for building AI agents in Databricks that coordinate multiple Genie spaces, automatically routing user questions to the appropriate data source without requiring users to know which space contains the answer. The demo shows creating a supervisor agent that bridges bakehouse and weather Genie spaces, handling multi-step queries and interpreting results for end users with minimal configuration.
Events[Demo] Introducing Agent Bricks: Auto-Optimized Agents Using Your Data
Agent Bricks demonstrates an auto-optimized agent that uses company data to generate a comprehensive launch report for a new product. The agent autonomously queries different internal agents (e.g., marketing, R&D, BizOps) to gather information on market trends, existing recipes, development timelines, costs, and more, culminating in a CEO-ready report.
Get Tuesday's version of this
Tracking Agent Bricks? The Tuesday email carries what moved across the whole ecosystem, not just this topic. Free, one-click unsubscribe.