Databricks

What does it take to move AI agents from prototype to production? Databricks' Amber R. and Craig Wiley walk through the full agent lifecycle, showing how to build, evaluate, deploy, and scale production-ready AI agents with Agent Bricks, AI Functions, and Genie Code. Watch this session, the keynotes, and more #DataAISummit 2026 content on demand → https://lnkd.in/g9WfYkfa

Connecting Amazon S3 data to Databricks now takes just a few clicks. The new external location creation flow makes one of the most important setup steps much easier, helping teams establish a secure, governed connection to S3 in minutes instead of working across multiple consoles. Less time configuring infrastructure. More time building with your data. https://lnkd.in/gGej5PzG

Anthropic's Claude Opus 5 is now available where your data already lives - governed, secure, and ready for custom AI applications and agents built with the data in your Lakehouse. On Databricks' internal coding bench, Opus 5 achieved SOTA performance. It delivers major gains in agentic coding, professional knowledge work, and long-horizon reasoning, with stronger performance per token across effort levels. Teams can build agents that handle complex workflows with fewer iterations and meaningfully lower cost, all governed by Unity AI Gateway Put Claude Opus 5 to work on your most important data, securely and at production scale. https://lnkd.in/g7ezUtq4

Magnite processes more than 2 trillion ad requests every day across video, display, and streaming platforms. To support that scale, the company adopted Databricks, Unity Catalog, and an Iceberg-first strategy to: ✓ Centralize governance across teams and systems ✓ Reduce consumer warehouse compute costs by 70% ✓ Cut average query times by 50–70% on a petabyte-scale log dataset Explore how Magnite is managing petabyte-scale ad data with better performance, lower costs, and consistent governance: https://lnkd.in/g53iB-QH

The organizations making progress with agentic AI are starting with the outcomes they want to drive, then building the data, governance, and application architecture to support them. In this eBook, Databricks executives and customers share practical guidance on outcome-first workflow design, semantic layers, agent lifecycle management, governance, and preserving choice as AI systems scale. They also outline five strategic plays for the next 18 months. Download your copy: https://lnkd.in/gf9PsBAy

What if the longer you drove your car, the better it got? Rivian’s Director of Big Data & AI, Michael Flynn, explains how the company streams tens of thousands of telemetry, sensor, and video data points from its vehicles into the lakehouse. Using machine learning and Genie, the team learns from that data and applies lessons from one vehicle across the fleet to improve the driving experience over time. Explore how Rivian uses data and AI to help its vehicles get better the more they’re driven.

We’re expanding our decade-long strategic partnership with Microsoft through the 2030s to help enterprises build AI grounded in their own business context. Our partnership is stronger than ever 🤝. Databricks is deepening its use of Azure Databricks and Azure Cobalt, while Microsoft continues to integrate the Databricks Data + AI Platform across its products. This means bringing capabilities like Genie directly into customer workflows through products like Microsoft 365, Teams and Copilot. Together, we’re helping customers scale enterprise AI with greater context, control, choice, cost efficiency, and reliability. https://lnkd.in/gmkxGxP5

What if the biggest barrier to useful enterprise AI is not intelligence, but context? Generic AI can handle broad productivity tasks, yet trusted data work depends on knowing which definitions, sources, and business rules matter. This guide explores that context gap and how Genie brings governed business data into the work teams already do. It also shows how analysts can shape that experience through three customer stories across retail, enterprise data, and supply chain. Download the guide → https://lnkd.in/gj2fR3Mh

Lakebase starts with a simple architectural shift: separate Postgres compute from the write-ahead log and data files that traditionally live on the same machine. Externalizing those components makes compute stateless and enables durable writes, elastic scaling, simpler high availability, and instant branching. That same design creates the foundation for LTAP. Operational data can be stored once in open columnar formats and read by both Postgres and Lakehouse engines, so analytics stays current without CDC pipelines, a second copy, or added load on transactional workloads. Databricks co-founder Reynold Xin explains how the architecture works and what it makes possible. https://lnkd.in/giscmddC

We're proud to be named a Leader in the 2026 Gartner® Magic Quadrant™: AI Platforms for Data Science and Machine Learning! Databricks is positioned highest in Ability to Execute and furthest in Completeness of Vision for the second year in a row. If you're evaluating AI platforms, this is the report to start with. Get the report: https://lnkd.in/gNG7vU_7