Databricks

We’re excited to announce a strategic partnership with Celonis to deliver a seamless pathway for Enterprise AI that continuously improves business operations. The integration connects the Databricks Platform with the Celonis Process Intelligence Platform, creating a continuous learning loop for intelligent action. The foundation is Delta Sharing, our open protocol that lets customers securely share live, governed data between both platforms without copying it. “Databricks brings secure, governed data and agentic development tools, and we bring the process intelligence AI needs to succeed. It’s a powerful combination” - Marc Kinast, VP Corporate Development at Celonis. Together, we’re helping organizations turn process insights into intelligent action: https://lnkd.in/eBcKeJ37

Enterprises can now securely connect AI agents to external tools and data with the new MCP Catalog and Agent Bricks! • MCP Catalog lets teams find and manage MCP servers in one place with Unity Catalog governance. • MCP in Databricks Marketplace expands your ecosystem with trusted partners like S&P Global, Moody's Corporation, Dun & Bradstreet, FactSet, Glean and Nasdaq. • Agent Bricks Multi-Agent Supervisor combines and creates powerful agents that reason over data and take action using servers from your MCP Catalog Together, they turn disconnected tools into secure, production-ready AI systems: https://lnkd.in/g5e_pc_2

What a keynote at #DataAIWorldTour Amsterdam! Eva Fors, Kevin Jonkergouw, Ari Kaplan, Kasey Uhlenhuth and Holly Smith set the stage, sharing how Dutch technology leaders can harness the potential of Agent Bricks and Lakebase. Wies Mensink showcased how Albert Heijn uses advanced analytics and AI on Databricks to optimize retail operations and elevate customer experiences. Teus Kappen highlighted how UMC Utrecht is transforming healthcare with data science and AI to support patient care and accelerate medical research. And SAP's Karl Fahrbach discussed with Samuel Bonamigo about how organizations can unify governance and data using SAP's Business Data Cloud. From retail to healthcare, organizations across the Netherlands are building trusted AI that accurately reasons with their data.

Thousands of data + AI projects have already been built with Databricks Free Edition, and we're excited to give our community the chance to showcase their work in the first Databricks Free Edition Hackathon! Build AI agents, train ML models, explore data engineering, create dashboards, or highlight other creative use cases — then share your project with our team. Hackathon submissions will be judged by Databricks co-founder, Reynold Xin, and staff. Submissions are now open through Nov. 14: https://lnkd.in/gsb9QwHY

Google's Gemini 2.5 models are generally available on Databricks! When speed and scale, and multimodal intelligence matter, Gemini models deliver — ideal for multimodal inputs, long-context RAG, summarization, and classification with fast, cost-efficient throughput. With AI Gateway and Agent Bricks, teams can unify data, governance, and observability while building agents that act on real business context. It’s now easier than ever to bring the power of Gemini models to trusted, production-ready AI systems. https://lnkd.in/gGKFvUiY

Enterprises can build intelligent AI agents powered by OpenAI GPT-5, now generally available on Databricks! GPT-5 shines even brighter when depth matters most — from complex reasoning and multi-turn agents to coding and tasks that demand nuance and high-quality judgment. With Agent Bricks and AI Gateway, teams can securely connect GPT-5 to governed enterprise data, build agents that reason with real context, and maintain unified governance and observability. GPT-5 on Databricks gives organizations the depth and reliability to power real-world, enterprise-scale AI. https://lnkd.in/gMDFBeUj

You can now build intelligent agents with every leading model on Databricks, powered by Agent Bricks. We know that flexibility fuels innovation, and that every model brings its unique strengths – whether that’s cost, performance or speed. OpenAI GPT-5 and Google's Gemini models now join Meta Llama and Claude as native models on Databricks. Build agents that combine your data with the best AI model for every task – with the freedom to switch as models continue to evolve at breakneck speed: https://lnkd.in/gxKdj2t7

Building accurate, trustworthy agents in Agent Bricks starts with better judges — because AI agents can only be as good as the ones evaluating them. We’re excited to streamline creation of domain-specific judges with new MLflow-powered capabilities in Agent Bricks: • Systematically align domain-specific evaluators with domain experts using new Tunable Judges • Describe evaluations in natural language with Agent-as-a-Judge • Easily create and maintain judges with Judge Builder, an intuitive visual workflow These updates make evaluation an engine for continuous improvement: https://lnkd.in/gXgQxU4c

🇩🇪 At #DataAIWorldTour Munich, 🇺🇦 Alexandra Klimova & Pieter Viljoen from Allianz shared how they’ve built a next-generation AI data platform with Databricks. Transforming a single team’s vision into a global enterprise solution. The Allianz Data Platform (ADP) unites governed data ingestion, ML microservices, and AI agent deployment, all while meeting complex regulatory requirements. This is trusted AI in action.

AI agents are shaking up every industry. The stakes are high when deploying agents into production, and the reality is that many projects stall in the pilot phase. The barrier to production: trust. This week, we’re announcing new Agent Bricks capabilities that make it easier to build agents that are accurate, governed, and production-ready. Agents you can trust. Here’s a look at what’s new 👇 https://lnkd.in/gC6_NxFh