What is AI-powered business intelligence?

Summary

• AI-powered business intelligence uses AI to make analytics self-service, so anyone in an organization can derive insights from data instead of relying solely on static reports built by data professionals.
• Databricks AI/BI is built on a compound AI system that combines multiple AI technologies, retrieval, ranking, and personalization to understand your organization's data semantics and usage patterns.
• It delivers two complementary experiences: AI/BI Dashboards, an AI-powered low-code dashboarding solution, and Genie, a conversational interface for natural language questions.
• Ask questions in everyday language. Genie lets business users query data without SQL and returns answers along with AI-generated explanations, drivers, and anomalies.
• Governed by Unity Catalog. AI/BI respects Unity Catalog permissions and lineage, and needs no separate data extract, so insights stay governed and fresh.

What is AI-powered business intelligence?

AI-powered business intelligence represents a shift in how organizations analyze data. Rather than relying solely on static reports and dashboards built by data professionals, AI-powered BI enables self-service analytics so that anyone in an organization can derive insights from data. Databricks AI/BI is built on a compound AI system, an architecture that combines multiple AI technologies, retrieval mechanisms, ranking systems, and personalization to understand organizational data semantics and usage patterns. It captures semantic understanding from interactions across the entire Databricks Platform, including ETL pipelines, lineage, and historical queries. See Databricks AI/BI and Introducing AI/BI: Intelligent Analytics for Real-World Data.

How Databricks AI/BI brings AI to business intelligence

Databricks AI/BI powers two complementary experiences:

• AI/BI Dashboards is an AI-powered, low-code dashboarding solution that includes conventional BI capabilities for answering a fixed set of business questions. Users describe what they want in natural language, and AI automatically generates dashboards using more than 20 visualization types, with cross-filtering and periodic PDF snapshots via email. Dashboards run natively on the platform and scale to large datasets without sampling or extract limits. See BI built for the AI era.
• Genie is a conversational interface that learns the underlying data and semantics continuously based on human feedback. It can answer a broad set of business questions through its reasoning capabilities, while still providing certified answers for query patterns specified by data teams. Genie responds with adaptable visualizations, seeks clarification when needed, and improves over time.

Natural language querying

Genie lets business users ask questions about their data in everyday language, without requiring SQL or technical knowledge. Users can ask questions such as "What's my pipeline by stage this quarter?" and receive not only answers but AI-generated explanations, drivers, anomalies, and relationships surfaced automatically. Genie supports follow-up questions and semantic understanding to help users uncover meaningful insights.

The compound AI system behind Genie

Genie operates as a compound AI system that integrates multiple interacting components rather than relying on a single AI model, combining natural language processing, machine learning models, retrieval systems, and ranking mechanisms. It draws on your full data lifecycle across Databricks, including ETL pipelines, lineage, and prior queries, to understand organizational context, and it generates SQL, charts, and summaries grounded in trusted data. Data teams can configure Genie with instructions, trusted assets, confidence voting, and quality monitoring to keep responses reliable and well governed. See How does AI/BI work?.

Governed by Unity Catalog

• Unified governance and lineage. AI/BI is integrated with Unity Catalog, so it adheres to global policies set by administrators, and lineage lets data producers trace how their data assets are used within AI/BI back to ingestion.
• Governed data access. Genie respects Unity Catalog permissions, so users only see data they are authorized to access.
• No separate extract. AI/BI works with your existing data, which removes the need to extract datasets into a separate BI engine, improving data freshness and simplifying governance.
• Unified semantics. Data teams can define consistent, reusable metrics through Unity Catalog semantics that Genie and dashboards use as a single source of truth.

Getting started

1. Review Databricks AI/BI to understand AI/BI Dashboards and Genie.
2. Create an AI/BI Dashboard by describing the visualizations you want in natural language.
3. Set up a Genie Agent over governed Unity Catalog data and let business users ask questions in everyday language.

FAQs

What is the difference between AI/BI Dashboards and Genie?

AI/BI Dashboards is a low-code way to build dashboards that answer a fixed set of business questions, while Genie is a conversational interface that answers a broader set of ad hoc questions through reasoning and improves over time from feedback.

Do business users need SQL to use AI-powered BI?

No. With Genie, users ask questions in everyday language and receive answers, visualizations, and AI-generated explanations without writing SQL.

How does AI-powered BI stay governed?

AI/BI is integrated with Unity Catalog, so it respects data permissions and lineage and uses governed semantics, and it works on your existing data without extracting it to a separate BI engine.