Databricks

Data sharing is a strategic pillar for data and AI innovations. With Delta Sharing, Databricks Marketplace and Clean Rooms, businesses can collaborate in real time while maintaining data privacy and governance. We recently unveiled new capabilities and enhancements to support organizations with their data sharing + democratization initiatives: - GA of Databricks Clean Rooms on Azure & AWS - New Delta Sharing improvements for secure data exchange - Integration of Partner Connect with Databricks Marketplace - SAP Databricks, powered by Delta Sharing - And more! Read the full roundup: https://dbricks.co/42kKPI3

As the creator of Minecraft, Mojang manages 130 million monthly players, generating vast data from player interactions and in-game activities. Analyzing this data across 22 platforms is crucial to maintaining an engaging experience. By leveraging Databricks Mosaic AI, Mojang: ? Delivers 2.5x more effective content recommendations in Minecraft Marketplace ? Analyzes player behavior in real-time to enhance game design and minimize friction ? Reduces data preparation time by 66%, enabling faster insights and innovation Discover how Mojang transforms complex player data into actionable insights using Databricks Data Intelligence Platform. https://dbricks.co/4lnYdD1

Databricks CEO Ali Ghodsi and Anthropic CEO Dario Amodei joined Bloomberg Television to discuss our latest partnership, which brings Claude access to the Data Intelligence Platform. “Both companies value things like safety, security, privacy — things that enterprise customers greatly value.” Catch the full conversation ?? https://dbricks.co/4iCa5PP

Join Ali Ghodsi, Co-founder and CEO of Databricks, and Dario Amodei, Co-founder and CEO of Anthropic, as they discuss the future of AI agents, best practices for building enterprise AI and why the best agents will use both general intelligence and data intelligence. Register here: https://dbricks.co/4iwJnZ2

Announcing native availability of Anthropic Claude 3.7 Sonnet in Databricks across AWS, Azure and GCP! Customers can now securely access Claude’s advanced reasoning, planning and agentic capabilities directly within Databricks, no setup required. Claude + the Databricks Data Intelligence Platform enables enterprises to build domain-specific AI agents on their unique data – with full governance and observability across the entire data and AI lifecycle. This launch marks the beginning of a strategic partnership between Anthropic and Databricks. We’re excited to see what our customers build and how they leverage domain-specific AI agents! https://dbricks.co/4hI3kdX

Our Chief AI Scientist, Jonathan Frankle, spoke to tons of our customers to understand the biggest challenges with enterprise AI. What he discovered: the problem is dirty data. He spoke to Will Knight about TAO (Test-time Adaptive Optimization), a new method – developed at Databricks – that combines reinforcement learning and synthetic data, making it possible to tune LLMs without labeled data. Read the full article via WIRED: https://dbricks.co/43pJzVd

Databricks has been named to the 2025 Enterprise Tech 30 list! ?? Selected by over 100 VCs and corporate development leaders, the #ET30 highlights the top private companies in enterprise technology — those with the most potential to change how tech enterprises operate. Thank you to Wing Venture Capital and Eric Newcomer for the recognition! https://dbricks.co/42gfKFj

Getting high-quality labeled data at scale is one of the biggest enterprise challenges with fine-tuning LLMs on specific tasks. TAO (Test-time Adaptive Optimization) is a new method that uses test-time compute and reinforcement learning to improve LLMs—no labeled data required. Requiring only example inputs, TAO often surpasses supervised fine-tuning with thousands of labeled data points and makes low-cost open source models outperform expensive proprietary ones. Learn more about this research from the Databricks Mosaic Research team. https://dbricks.co/420Xte1

The future of AI is domain-specific agents. Yet most enterprises still struggle to build production-ready AI applications. Join this virtual event to hear from AI experts as they share how Databricks applies data intelligence to AI agents – plus real-world examples of companies solving domain-specific challenges. Register here: https://dbricks.co/4iwJnZ2

Keep your AI agents online—no matter what. Dependency on a specific foundation model provider increases the risk when that specific service goes down unexpectedly, causing disruptions and downtime for AI-driven applications. Mosaic AI Gateway’s traffic fallbacks ensure seamless failovers across multiple providers, regions, and resources—so your AI agents stay operational, even during outages or traffic spikes. Customers like Erste Group rely on AI Gateway to keep AI operations resilient and uninterrupted. https://dbricks.co/3DvPWM4