Databricks AI-BI Genie Conversational Analytics Cheat Sheet

Updated 2026-05-23

Next Topic: Databricks Dashboards Cheat Sheet

🧠Study flashcards on this topic72 cards · spaced repetition→

Databricks AI/BI Genie is a conversational analytics interface that lets business users ask questions in plain English and receive answers backed by SQL queries on Unity Catalog data. Unlike simple chatbots, Genie uses a compound AI system — combining language models, semantic metadata, prompt matching, and query verification — to generate and validate SQL against your actual data. It became generally available (GA) on June 12, 2025, and runs entirely within the Databricks platform governance model, meaning Unity Catalog row filters and column masks apply automatically to every query Genie generates.

What This Cheat Sheet Covers

This topic spans 14 focused tables and 105 indexed concepts, 72 flashcards. Below is a complete table-by-table outline of this topic, spanning foundational concepts through advanced details.

Table 1: Core Concepts and ArchitectureTable 2: Setting Up a Genie SpaceTable 3: Permissions and Access ControlTable 4: Knowledge Store — General Instructions and MetadataTable 5: Knowledge Store — SQL ExpressionsTable 6: Knowledge Store — Trusted Assets and Example QueriesTable 7: Prompt Matching — Format and Entity AssistanceTable 8: Benchmarks and EvaluationTable 9: Genie API and Python SDKTable 10: Conversation API PatternsTable 11: AI/BI Dashboard IntegrationTable 12: Unity Catalog Metric Views (Semantic Layer)Table 13: Genie Agent Mode and Advanced FeaturesTable 14: Limits, Troubleshooting, and Best Practices

Quick IndexSubscribe to unlock

A jump-to index of every table row in this cheat sheet.

Mind MapSubscribe to unlock

An interactive map of every table and concept in this topic.

Table 1: Core Concepts and Architecture

Genie is not a single model but a system of cooperating components. Understanding its architecture helps you tune it correctly and set realistic expectations about what it can and cannot answer accurately without additional configuration.

Concept	Example	Description
Genie Space	A "Sales Analytics" space connected to `sales.crm.opportunity` and `sales.crm.accounts`	A curated conversational interface built by domain experts (analysts/data engineers) that exposes specific Unity Catalog tables and contextual instructions to business users.
Compound AI System	NL → SQL generation → Inspect verification → result	• Genie is not a single LLM • it chains multiple AI steps — query generation, verification, result summarization — into a coordinated pipeline
Knowledge Store	Table descriptions + SQL expressions + example queries	• The living semantic model attached to a Genie space • curators add metadata, instructions, and SQL snippets that teach Genie domain meaning
Unity Catalog Requirement	Tables must be registered in Unity Catalog to be added to a space	• Genie only operates on Unity Catalog tables and views • all access is governed by existing UC row filters, column masks, and privileges
Trusted Response	UI badge "Trusted" on an answer	• Response label shown when Genie's generated SQL exactly matches a pre-verified trusted asset (example query or UC SQL function) • signals highest confidence
Inspect Feature	User clicks "Inspect" on a Genie response	Advanced reasoning step that reviews and validates the generated SQL via sub-query decomposition to improve accuracy before results are returned.
Agent Mode / Deep Research	"Why did revenue drop in Q3?" triggers multi-step investigation	• Multi-step agentic reasoning for complex "why/how" questions • Genie forms hypotheses, runs multiple queries, and synthesizes an explanation
Companion Genie Space	AI/BI Dashboard automatically gets an "Ask Genie" button	• Auto-generated Genie space attached to an AI/BI Dashboard • inherits the dashboard's datasets (up to 30) so users can ask follow-up questions conversationally

Table 2: Setting Up a Genie Space

Creating a well-configured space is the foundation of a successful Genie deployment. The setup determines which data users can query and how much context Genie has to interpret their questions accurately.

Concept	Example	Description
Create Space	Databricks UI → New → Genie Space	• Spaces are created in the Databricks workspace UI • requires a SQL warehouse and at least one Unity Catalog table or view

The rest of this section is part of the full cheat sheet

Table 3: Permissions and Access Control

Genie inherits the full Unity Catalog permission model and adds its own space-level ACLs. Understanding both layers prevents common "user can't see data" issues while keeping data governance intact.

Concept	Example	Description
CAN MANAGE	Space owner / workspace admin	Full control: edit settings, knowledge store, permissions, and delete the space.

The rest of this section is part of the full cheat sheet

Table 4: Knowledge Store — General Instructions and Metadata

The quality of a Genie space is almost entirely determined by the richness of its Knowledge Store. General Instructions provide global context, while table/column metadata provides structural understanding.

Concept	Example	Description
General Instructions	`"The 'stage' column uses Salesforce stage names: 'Closed Won', 'Closed Lost', 'Prospecting'."`	• Free-text instructions applied to every question • use for business rules, terminology definitions, timezone conventions, and table relationship hints

The rest of this section is part of the full cheat sheet

Table 1: Core Concepts and Architecture

Concept	Example	Description
Genie Space	A "Sales Analytics" space connected to `sales.crm.opportunity` and `sales.crm.accounts`	A curated conversational interface built by domain experts (analysts/data engineers) that exposes specific Unity Catalog tables and contextual instructions to business users.
Compound AI System	NL → SQL generation → Inspect verification → result	• Genie is not a single LLM • it chains multiple AI steps — query generation, verification, result summarization — into a coordinated pipeline
Knowledge Store	Table descriptions + SQL expressions + example queries	• The living semantic model attached to a Genie space • curators add metadata, instructions, and SQL snippets that teach Genie domain meaning
Unity Catalog Requirement	Tables must be registered in Unity Catalog to be added to a space	• Genie only operates on Unity Catalog tables and views • all access is governed by existing UC row filters, column masks, and privileges
Trusted Response	UI badge "Trusted" on an answer	• Response label shown when Genie's generated SQL exactly matches a pre-verified trusted asset (example query or UC SQL function) • signals highest confidence
Inspect Feature	User clicks "Inspect" on a Genie response	Advanced reasoning step that reviews and validates the generated SQL via sub-query decomposition to improve accuracy before results are returned.
Agent Mode / Deep Research	"Why did revenue drop in Q3?" triggers multi-step investigation	• Multi-step agentic reasoning for complex "why/how" questions • Genie forms hypotheses, runs multiple queries, and synthesizes an explanation
Companion Genie Space	AI/BI Dashboard automatically gets an "Ask Genie" button	• Auto-generated Genie space attached to an AI/BI Dashboard • inherits the dashboard's datasets (up to 30) so users can ask follow-up questions conversationally

Table 2: Setting Up a Genie Space

Concept	Example	Description
Create Space	Databricks UI → New → Genie Space	• Spaces are created in the Databricks workspace UI • requires a SQL warehouse and at least one Unity Catalog table or view

Table 3: Permissions and Access Control

Genie inherits the full Unity Catalog permission model and adds its own space-level ACLs. Understanding both layers prevents common "user can't see data" issues while keeping data governance intact.

Concept	Example	Description
CAN MANAGE	Space owner / workspace admin	Full control: edit settings, knowledge store, permissions, and delete the space.

Table 4: Knowledge Store — General Instructions and Metadata

Concept	Example	Description
General Instructions	`"The 'stage' column uses Salesforce stage names: 'Closed Won', 'Closed Lost', 'Prospecting'."`	• Free-text instructions applied to every question • use for business rules, terminology definitions, timezone conventions, and table relationship hints

Databricks AI/BI Genie Conversational Analytics Cheat Sheet

Table 1: Core Concepts and Architecture

Table 2: Setting Up a Genie Space

Table 3: Permissions and Access Control

Table 4: Knowledge Store — General Instructions and Metadata

Databricks AI/BI Genie Conversational Analytics Cheat Sheet

Table 1: Core Concepts and Architecture

Table 2: Setting Up a Genie Space

Table 3: Permissions and Access Control

Table 4: Knowledge Store — General Instructions and Metadata