What’s New in the AI Platform: Agents for ML Engineering, Our Deep Learning Platform, and New Capabilities for Real-Time ML
Summary
Databricks shipped Genie Code, a coding agent for ML engineering, and AI Runtime, a serverless GPU platform for deep learning. Power real-time ML at scale with new Feature Store and Model Serving capabilities, including streaming features and high-QPS serving.
Summary generated by brickster.ai. For the full article, follow the source link above.
More from Databricks Blog
The Partner Well-Architected Framework: What's New and What's Next
The Partner Well-Architected Framework (PWAF) now includes the AI Partner Dev Kit, expanded architecture patterns, and the Firefly reference application to accelerate development and simplify integrations for partners. These new capabilities provide AI-ready architecture guidance and best practices for building on or connecting to Databricks.
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.
Becoming the most comprehensive data & AI ecosystem on earth
Databricks is expanding its Data and AI partner ecosystem with new Marketplace, Apps, OpenSharing, and Genie Agent capabilities for building, distributing, sharing, and monetizing solutions. These updates accelerate deal velocity, provide access to 20,000+ Databricks customers, and unlock new AI-era business models like pay-per-question.
Design Beautiful Dashboards in AI/BI
Databricks AI/BI now offers dashboard themes to customize fonts, colors, and visualization palettes for on-brand dashboards. Apply workspace themes once to automatically carry your theme across all dashboards.
What’s new in Genie Code at Data + AI Summit 2026
Genie Code now features a full-page command center for managing complex, multi-threaded data and ML work, providing thread status, review points, and quick access to instructions, skills, and connectors. Scheduled tasks and Genie ZeroOps are coming soon to enable autonomous execution and extend automation into data and ML operations.
What’s new in Databricks Platform security and compliance at Data + AI Summit 2026
Automatic Identity Management (AIM) for Entra ID is now GA on AWS and GCP, with AIM for Okta in Public Preview, alongside new Context-Based Ingress policies and expanded Private Link support for Lakebase and account-level services. Databricks also announced new certifications, regional compliance programs, broader AWS GovCloud support for AI services, and upcoming FedRAMP High support on Azure Commercial.