Home / Data Analyst practice test / Developing, Sharing, and Maintaining AI/BI Genie spaces

Free · 7 questions with explanations

Developing, Sharing, and Maintaining AI/BI Genie spaces: Databricks Data Analyst Associate Practice Questions

Exam-style questions on Developing, Sharing, and Maintaining AI/BI Genie spaces. Pick your answer, then open the explanation to see why it's right — and why the other options are wrong.

Take a full timed test free → All Data Analyst topics
1 Developing, Sharing, and Maintaining AI/BI Genie spaces

A Genie space author has added several example SQL queries to the space's instruction context. Under what precise condition does Genie label a response as 'Trusted' when using an example SQL query?

  1. AOnly when a parameterized example SQL query is used to generate the response — non-parameterized (static) example queries provide context to Genie but are never labeled Trusted, even when directly applied
  2. BWhenever any example SQL query from the space's instruction context is referenced during response generation, regardless of whether it is parameterized or static, as long as the query runs without errors on the warehouse
  3. COnly when the user explicitly clicks a 'Request trusted answer' button before submitting their question, which instructs Genie to restrict its SQL generation to only the pre-approved example queries added by the space author
  4. DWhen the generated SQL response has been manually reviewed and approved by a user with at least CAN MANAGE permission on the Genie space using the approve workflow in the Monitoring tab
  5. EWhenever the user's question text matches any saved sample question in the space settings with an exact string match, which triggers Genie to return the associated example query result with a Trusted label
Show answer & explanation

Correct answer: A

WHY A: Per Databricks documentation, a response is labeled Trusted only when the exact text of a parameterized example SQL query is used OR when a Unity Catalog SQL function is invoked. Static (non-parameterized) example queries guide Genie in generating SQL but are explicitly NOT labeled as Trusted even when referenced. WHY NOT B: Static example queries do not trigger the Trusted label even if they run without errors; the parameterization requirement is essential — it ensures the query logic is deterministic and the output is verified. WHY NOT C: There is no 'Request trusted answer' button; the Trusted designation is automatic and is determined by whether a parameterized example query or SQL function was used — not by user request. WHY NOT D: The Monitoring tab allows authors to review and correct responses, but manually approving a response there does not retroactively apply a Trusted label to it. WHY NOT E: Sample questions in settings are navigational hints shown in the chat UI for user discovery; an exact match against a sample question does not independently trigger the Trusted label — the trusted asset mechanism is SQL-level, not question-title-level.

2 Developing, Sharing, and Maintaining AI/BI Genie spaces

A technical architect asks how Genie generates SQL responses differently from a conventional single large language model (LLM) chatbot. Which statement accurately describes Genie's underlying AI architecture?

  1. AGenie uses a compound AI system — a combination of multiple interacting components working together — rather than relying on a single large language model to translate natural language questions into SQL queries
  2. BGenie uses a single large language model that has been fine-tuned exclusively on Databricks SQL query syntax, enabling it to generate accurate SQL without requiring any additional system-level components or orchestration logic
  3. CGenie routes every natural language question to a rules-based SQL template engine that matches keyword patterns against pre-registered query blueprints, falling back to an LLM only when no template matches
  4. DGenie applies a traditional keyword-extraction pipeline that tokenizes the user's question, identifies column names via fuzzy matching, and then constructs SQL directly from the matched column list without invoking any language model
  5. EGenie sends each user question to multiple competing large language models in parallel, scores each generated SQL for syntactic validity, and returns the result from whichever model achieves the highest validation score
Show answer & explanation

Correct answer: A

WHY A: According to Databricks documentation, Genie uses a compound AI system — a design that combines multiple interacting components instead of a single LLM — to interpret business questions and generate answers. This architecture provides better performance and flexibility than a single-model approach. WHY NOT B: Genie is not based on a single fine-tuned LLM; the compound AI design is explicitly intended to go beyond the limitations of any single model. WHY NOT C: Genie is not a rules-based template engine; it leverages generative AI guided by metadata, instructions, and knowledge store context to produce dynamic SQL. WHY NOT D: Keyword-extraction and fuzzy-matching pipelines describe traditional NLP, not Genie's generative AI architecture. Genie translates natural language to SQL semantically, not through keyword templates. WHY NOT E: Genie does not run multiple competing LLMs in parallel and select by validation score; the compound AI system uses coordinated sequential components, not a competition model among independent LLMs.

3 Developing, Sharing, and Maintaining AI/BI Genie spaces

A data engineer is setting up a Genie space for a large enterprise and wants to include as many data assets as possible. What is the maximum number of Unity Catalog tables or views that can be added to a single Genie space?

  1. A10 tables or views — Genie spaces are designed for focused analytical domains, and performance degrades significantly beyond 10 assets due to context window limitations in the LLM integration layer
  2. B50 tables or views — this is the hard limit per space, and any attempt to exceed it results in an error during space configuration until the author removes excess tables from the asset list
  3. C100 tables or views — matching the overall instruction limit in a Genie space so that the total number of tables never exceeds the combined instruction count capacity of a single space
  4. D30 tables or views — Genie spaces support adding up to 30 Unity Catalog tables or views, including managed tables, external tables, views, metric views, and materialized views
  5. EUnlimited — there is no documented cap on the number of Unity Catalog data assets that can be added to a Genie space, provided all assets are registered in a Unity Catalog-enabled workspace
Show answer & explanation

Correct answer: D

WHY D: Databricks documentation specifies that a Genie space can include up to 30 Unity Catalog tables or views, covering managed tables, external tables, foreign tables, views, metric views, and materialized views. WHY NOT A: The limit is 30, not 10. The documentation does not describe a performance-degradation threshold at 10 assets. WHY NOT B: The limit is 30, not 50. There is no documented 50-table limit for Genie spaces. WHY NOT C: The 100-instruction limit and the 30-table limit are separate constraints that do not mirror each other; each example SQL query, SQL function, and the general instructions text block together count toward the 100 instructions; table count is tracked separately. WHY NOT E: There is a documented limit of 30 data objects; it is not unlimited.

4 Developing, Sharing, and Maintaining AI/BI Genie spaces

After a Genie space is configured and shared with a business team, the author wants to distribute the space via a URL so that team members can access it directly from a bookmark or internal wiki. Which UI action provides the shareable link to the Genie space?

  1. AClick Configure > Settings and then select 'Generate public embed code', which creates a permanent embeddable iframe URL that can be pasted into any wiki or internal portal for direct access
  2. BOpen the Monitoring tab in the Genie space, locate the 'Space URL' field at the top of the monitoring dashboard, and copy the auto-generated access URL for distribution to business users
  3. CNavigate to the workspace Object Browser, right-click the Genie space object, choose 'Get shareable link', and then copy the workspace-relative URL generated for the space in the dialog box
  4. DOpen the Genie space, click the kebab menu in the upper-right corner, choose 'Export link', and then select either 'Workspace link' or 'Account link' depending on whether the recipients have workspace access
  5. EOpen the Share dialog by clicking Share on the Genie space, and then click the 'Copy link' button at the bottom of the dialog to obtain a shareable URL that privileged users can use to open the space
Show answer & explanation

Correct answer: E

WHY E: According to Databricks documentation, to share a Genie space via URL the author opens the Share dialog (by clicking the Share button), enters users or groups, and then clicks the 'Copy link' button at the bottom of the Share dialog to get a shareable link. This link allows privileged users to open the Genie space directly. WHY NOT A: There is no 'Generate public embed code' option under Configure > Settings; embedding Genie into external apps uses the Genie API, not an auto-generated iframe constructor from the settings panel. WHY NOT B: The Monitoring tab shows questions, responses, and feedback; it does not contain a 'Space URL' field for generating access links. WHY NOT C: The workspace Object Browser does not have a right-click 'Get shareable link' option for Genie spaces; link sharing is done through the Share dialog within the space itself. WHY NOT D: The kebab menu in the upper-right provides options such as Clone and settings navigation, not an 'Export link' option with workspace/account link variants.

5 Developing, Sharing, and Maintaining AI/BI Genie spaces

After a Genie space has been deployed, business users mark several responses as incorrect using the 'Fix it' feedback option. The author notices the patterns in the Monitoring tab. What must the author do to actually improve Genie's responses to those types of questions?

  1. AManually update the space's instructions, example SQL queries, or trusted assets based on patterns observed in Monitoring — Genie does not automatically improve from user feedback ratings alone
  2. BRe-publish the Genie space by clicking 'Refresh and Retrain' in the Configure menu, which triggers Databricks to automatically absorb all flagged feedback responses and fine-tune the underlying language model for the space
  3. CAsk users to flag the problematic questions as 'Request review' and then mark them as confirmed correct in the Monitoring tab, which retroactively teaches Genie the correct SQL pattern for those question categories
  4. DContact Databricks support to submit the flagged interactions as model improvement tickets, which are periodically reviewed and incorporated into Genie's model weights in subsequent platform release updates
  5. EDelete the problematic example SQL queries from the space's instructions, restart the associated SQL warehouse, and then allow Genie to regenerate fresh instructions based solely on Unity Catalog metadata without prior context
Show answer & explanation

Correct answer: A

WHY A: Databricks documentation explicitly states that Genie's behavior does not change based on user feedback alone. Authors must use the feedback patterns they observe in the Monitoring tab to manually update instructions, example SQL queries, or trusted assets. The Monitoring tab identifies problem areas; actual improvement requires the author to apply changes. WHY NOT B: There is no 'Refresh and Retrain' option in the Configure menu; Genie does not fine-tune an LLM from user feedback within the space. Improvement requires author-driven instruction updates, not automated retraining. WHY NOT C: Marking responses as confirmed correct in the Monitoring tab does update the review status for business users, but it does not automatically teach Genie new SQL patterns; the author still must add or update instructions manually. WHY NOT D: Individual workspace-level user feedback is not submitted to Databricks as model improvement tickets; Genie space optimization is handled by the space author through configuration changes, not Databricks engineers. WHY NOT E: Deleting example queries and restarting the warehouse reduces the contextual guidance that helps Genie generate accurate SQL; this would likely worsen performance, not improve it. The correct action is to add or refine instructions for the problematic question categories.

6 Developing, Sharing, and Maintaining AI/BI Genie spaces

A data analyst wants to add a complex SQL function to a Genie space so that Genie can invoke it as a verified trusted asset when answering specific business questions. Where must this function be registered before it can be added to the Genie space?

  1. AIt must be saved as a .sql file in the Databricks workspace file browser under the author's user folder, and then referenced by relative path from within the Genie space's SQL Queries configuration tab
  2. BIt must be created inside the Genie space editor itself using the Built-in Function Builder, which compiles and sandboxes the function logic within the Genie space's private execution context
  3. CIt must be uploaded as a Python UDF jar into the SQL warehouse's external libraries configuration so that Genie can dynamically load and invoke it at query runtime through the warehouse compute layer
  4. DIt must be hosted in a Databricks Repos Git-tracked SQL file and linked to the Genie space via its repository URL, allowing Genie to pull and compile the function definition at each invocation
  5. EIt must be registered as a Unity Catalog SQL table function first; only after that registration can it be added to a Genie space as a trusted asset under Configure > Instructions > SQL Queries
Show answer & explanation

Correct answer: E

WHY E: Databricks documentation explicitly states that SQL functions used as trusted assets in Genie must be registered in Unity Catalog. The function is a Unity Catalog securable object, and once registered, a user with at least CAN EDIT permission on the Genie space can add it via Configure > Instructions > SQL Queries. WHY NOT A: Saving a .sql file in the workspace file browser does not register it as a Unity Catalog function; Genie cannot reference workspace-path SQL files as callable trusted assets. WHY NOT B: There is no Built-in Function Builder inside the Genie space editor; functions must be created externally and then registered in Unity Catalog before being added to the space. WHY NOT C: Python UDF jars in library configurations are not the mechanism for Genie trusted assets; Genie trusted asset functions must be SQL table functions registered in Unity Catalog. WHY NOT D: Linking a Repos SQL file by URL does not make it a callable Unity Catalog function; Unity Catalog registration is required specifically so the function is a governed, executable object with proper permissions.

7 Developing, Sharing, and Maintaining AI/BI Genie spaces

A data analyst creates a new AI/BI Genie space and then immediately shares it with a business analytics group. What permission level is the space creator automatically assigned on the space at the moment of creation?

  1. ACAN VIEW — the creator begins with read-only access to prevent accidental changes during the initial sharing phase and must explicitly self-upgrade to a higher permission level through the workspace admin console
  2. BCAN EDIT — creators are assigned the intermediate permission that allows them to configure the space and update instructions but requires a workspace admin to promote them to CAN MANAGE for ownership tasks
  3. CCAN RUN — this is the baseline permission that all users including the creator receive upon creation, and elevated permissions must be granted by the workspace admin after the space is reviewed and approved
  4. DCAN MANAGE — Genie space creators are automatically assigned CAN MANAGE permissions on any Genie space they create, granting them full control including editing, sharing, monitoring, and deleting the space
  5. EOwner — creators are designated as the exclusive Owner of the space, a role that supersedes the standard ACL permission tiers and can only be transferred through the workspace admin account settings panel
Show answer & explanation

Correct answer: D

WHY D: Databricks documentation explicitly states that Genie space creators automatically have CAN MANAGE permissions on spaces they create. CAN MANAGE is the highest permission tier, granting full control over configuration, sharing, monitoring, and deletion. WHY NOT A: CAN VIEW is a read-only tier; assigning a creator only CAN VIEW would prevent them from configuring or editing their own space, which contradicts the documented behavior. WHY NOT B: CAN EDIT allows configuration updates but does not grant the full management capabilities (like transferring ownership or seeing the monitoring page at the CAN MANAGE level); the creator receives the higher CAN MANAGE level by default. WHY NOT C: CAN RUN is the minimum access level for business users to interact with the space; the creator receives CAN MANAGE, not CAN RUN, upon creation. WHY NOT E: There is no separate 'Owner' role that supersedes the CAN MANAGE tier; CAN MANAGE is the top-level Genie space ACL permission, and it is what creators receive automatically.

Take the full Data Analyst practice test →

Keep going

Other Data Analyst topics


Executing queries using Databricks SQL and Databricks SQL Warehouses 8 practice questions → Working with Dashboards and Visualizations in Databricks 8 practice questions → Analyzing Queries 8 practice questions → Managing Data 5 practice questions → Securing Data 5 practice questions → Understanding of Databricks Data Intelligence Platform 5 practice questions → Data Modeling with Databricks SQL 4 practice questions → Importing Data 4 practice questions →