The 3 questions to answer to take AI from experimentation to impact
Excerpt
Companies are starting to see the potential of AI in their businesses. Today, 60%...
Excerpt from the source feed. For the full article, follow the source link above.
More from Databricks Blog
Inside the infrastructure strategies propelling AI leaders
AI adoption is starting to translate into real-world returns. But as efforts accelerate,...
How we keep GPUs reliable across Databricks AI
Databricks AI uses a multi-pronged approach to ensure GPU reliability, addressing crashed jobs, silent slowdowns, and numerical corruption through pre-workload validation, in-load monitoring, and inter-node fabric health checks. This system, stress-tested by diverse, large-scale workloads like RL for agentic coding, catches issues like fabric flakiness and thermal hotspots before they impact broader production.
Celebrating the Winners of the 2026 Built-On Databricks Startup Challenge
Databricks is celebrating the winners of the 2026 Built-On Databricks Startup Challenge, a global competition for early-stage B2B startups building core products on the Databricks platform. VisionHeight took the grand prize for its attack intelligence infrastructure, an innovation in agentic threat intelligence, alongside other winners in areas like web search and retail.
Granular Usage Attribution for dbt Pipelines with Query Tags
Databricks now supports granular usage attribution for dbt pipelines using query tags, allowing you to track costs and compute time by team, cost center, project, and environment without modifying SQL models. A reference project with a dbt pipeline, analytics dashboard, and scheduled job is available via Declarative Automation Bundles for easy deployment.
Beyond dashboards: Introducing Decision Execution Platforms
Databricks FDE introduces Decision Execution Platforms (DEPs), a new analytics category that automates the full executive decision loop—signal, decision, execution, and outcome—on your governed Databricks infrastructure. Unlike traditional BI, DEPs turn insights into measured action with a governed Decision Log, as demonstrated by an early Fortune 100 retail deployment targeting a $100M annual fulfillment gap.
Forecasting at the speed of modern retail
MMF Agent, a new guided AI workflow built on Genie Code, makes Databricks' multi-model forecasting framework accessible without deep data science expertise. Databricks practitioners can now compress setup time, clean training data, and achieve multi-model accuracy improvements previously limited to specialist teams.