Sign up

Practical Databricks expertise for engineers building on the lakehouse and putting agents to work

Databricks in a real Microsoft estate

Azure Databricks is a first-party Azure service, and many SQLBits attendees run it alongside Fabric, SQL Server and Power BI. Learn how the pieces fit, where to draw the lines, and how teams are making a mixed estate work. 

Practical Databricks skills you can put to work

Lakeflow pipelines, Unity Catalog governance, Genie and Agent Bricks, cost control, CI/CD with Asset Bundles and migrating from SQL Server and Synapse. Sessions are practical and vendor-honest. 

Learn from people who ship on Databricks

Databricks specialists, MVPs and consultants who build on the platform every day will share what works, what doesn’t and what they would do differently. 

Because your SQLBits pass includes virtual, you don't have to choose between sessions. Watch sessions live, catch up on anything you miss and revisit the content afterwards when you need it.

The challenges we’ll help you solve 

If any of these sound familiar, you will find people at SQLBits who have already been there. 

  • We can’t predict or attribute DBU spend across teams, serverless and AI workloads. 
  • Serverless, classic or SQL warehouse: which compute is cheapest for each workload? 
  • Hive metastore to Unity Catalog migration is still unfinished. 
  • Governance stops at the platform edge, but we also have Fabric, Snowflake and external AI tools. 
  • Databricks, Fabric or both? Nobody agrees who owns what in the same estate. 
  • Medallion layers multiply into hundreds of near-duplicate tables nobody owns. 
  • CI/CD is inconsistent: Asset Bundles, GitHub Actions and Power BI deployment are rarely in one pipeline. 
  • Our SQL Server and Synapse estate needs moving, but T-SQL, SSIS and stored procedures will not convert by themselves. 
  • Agents look good in demos but are unpredictable in production. How do we evaluate, trace and trust them? 
  • The platform changes monthly, and models, endpoints and APIs get retired. 
  • Our team knows SQL and BI. Spark, Python, DevOps and agents are the gap. 
  • Delta or Iceberg, and can our data leave cleanly if we need it to? 

Using Databricks to strengthen your company’s AI capability

Build and ship agents on governed data

Agent Bricks adds managed memory, sandboxed code execution and deployment of LangGraph, CrewAI and similar agents on Databricks Apps. 

Document intelligence in SQL

ai_extract and ai_classify (GA, June 2026) and ai_query (GA) let analysts use AI without writing Python. 

Search and retrieval

Vector Search was renamed AI Search in June 2026 and now includes full-text indexes without embeddings. 

Operational Postgres next to the lakehouse

Lakebase, serverless Postgres with autoscaling, went GA on 22 Jan 2026 per Databricks documentation. 

Governance for models, agents and MCPs

Unity Catalog permissions for foundation models are GA, with attribute-based access control for models in beta. 

Observability and evaluation

MLflow trace storage in Unity Catalog is GA (July 2026), keeping agent traces and monitoring in the lakehouse. 

Interoperability with Fabric, OneLake and Azure

OneLake catalog federation is GA (June 2026), and publishing a Unity Catalog catalog to Fabric as a read-only mirrored catalog is in public preview. 

Where agentic AI can help

Eight places agents are already earning their keep in Databricks, and where SQLBits sessions will show you how. 

Writing and fixing code in notebooks and SQL

Genie Code (formerly Databricks Assistant) plans and runs multi-step data engineering, data science and dashboard tasks. Agent mode went GA in March 2026. 

Self-service questions for business users

Genie One answers questions in plain English over governed data and can act via MCP tools. Reported GA in June 2026, with the Genie One MCP server GA on 25 Sept 2026. 

Multi-agent orchestration

Agent Bricks Supervisor Agent (GA, 10 Feb 2026) routes questions across Genie spaces, document agents and MCP tools using the asking user’s permissions. 

Pipeline authoring

Lakeflow Designer (GA, 16 June 2026) is an AI-assisted drag-and-drop canvas that generates pipelines. 

Pipeline and model operations

Genie ZeroOps (preview) finds root causes, tests a fix in a clone and waits for human approval. 

Governed tool access

Agents reach external tools through MCP, governed by Unity Catalog, including Managed OAuth for external MCP servers. 

Cost and performance control

Query tags (GA, 30 Sept 2026), query history in system tables and Budgets (GA, July 2026) let teams attribute spend. Genie Code scheduled tasks can run recurring checks unattended. 

Migrations

Lakebridge (free, open source) and the assistant’s /migrate command translate T-SQL, Teradata, Oracle and Snowflake SQL to Databricks SQL. An agentic code converter is in beta. 

Key Speakers

All of these speakers have appeared on the SQLBits stage – check our 2027 agenda when it’s announced for details of our 2027 Databricks sessions and speakers. 

Simon Whiteley, CTO, Advancing Analytics

Databricks MVP and Microsoft Data Platform MVP, creator of the Advancing Spark YouTube series and a Data + AI Summit speaker. 

 

Holly Smith, Developer Advocate, Databricks

One of Databricks’ earliest team members, with 10+ years in data and AI and a passion for teaching data engineering. 

 

Gavin Campbell, Data engineering practitioner

A long-time SQL Server and BI professional turned Databricks DevOps practitioner. 

 

Gavita Regunath, Chief AI Officer, Advancing Analytics

Microsoft AI MVP and Databricks MVP, and a Data + AI Summit 2025 session host. 

Check out some of our recent Databricks sessions

AgentBricks vs Mosaic AI
Databricks now offers two paths for building enterprise AI agents: AgentBricks and Mosaic AI. Join Tori and Gavi, 2 AI MVPs as they compare these capabilities, architecture, and ideal use cases. We’ll explore how AgentBricks simplifies agent creation with automated evaluation, synthetic data, and low‑code workflows, while Mosaic AI provides the broader infrastructure for model governance, routing, observability, and production‑grade agent systems. Attendees will learn when to use each, how they work together, and how to build scalable agentic applications across the Databricks platform.
 
An Apache Spark query's journey through the layers of Microsoft Fabric
Follow an Apache Spark SQL query from optimizer to execution across Spark and Microsoft Fabric. Catalyst, Tungsten, Fabric’s native C++ engine, and the Delta Lake/Parquet storage layers are covered, grounded in Spark source code, with practical tips to optimize for speed and cost.
 
Databricks Asset Bundles: Putting the Engineering into Data Engineering
Learn how to automate Databricks deployments with Asset Bundles and GitHub actions
 
Databricks vs Fabric: When to choose, when to combine, and why it’s confusing
Join this session to move beyond opinions and slide decks. Through practical demos, you’ll see how both platforms behave when solving the same real-world problems, and where the differences actually matter in delivery.
 
Fabric + Databricks + Power BI: Making the Integration Actually Work
Everyone talks about Fabric and Databricks together, few make it work in production. This session shares real-world patterns for integrating Databricks, Fabric and Power BI, covering semantics, performance, CI/CD, governance, and cost trade-offs beyond the PoC.
 
Medallion architecture
This session discusses the Medallion architecture of modern data platforms. You will learn what the Medallion architecture is, and you will also learn about its variations and why they might make sense.
 
No Code Agents with AgentBricks
AgentBricks is a new framework for designing, testing, and deploying production-grade AI agents natively on the Databricks Lakehouse. This session introduces the architecture behind AgentBricks and shows how it enables teams to create reliable, auditable, and enterprise-ready agents. We’ll explore how AgentBricks leverages Unity Catalog governance, Lakehouse data pipelines, and Model Serving to orchestrate multi-step agent workflows with full observability.
 
Semantic Bridge: Importing Databricks Metric Views into Tabular semantic models
Use Databricks Metric Views and Fabric Power BI semantic models interchangeably with the Semantic Bridge. This new feature from Tabular Editor allows you to seamlessly import the full structure of a Metric View into a Tabular semantic model.
 
Unlock Unity Catalog: Practical Strategies for Secure, Scalable Data Management
Learn to organize Unity Catalog resources for secure, scalable environments. Get tips on structuring catalogs, managing permissions, handling PII, and applying privacy controls—ideal for data pros seeking actionable data management insights.
 

Take a tour of SQLBits 2025

Bring your Databricks questions to the people who can answer them. Book your SQLBits 2027 pass today.