Databricks-GenAI-Assoc Practice Test 15 — 15 Questions

Question 1

A team is designing a Databricks RAG application that must return answers with citations to source documents. They plan to use Databricks Vector Search and want the LLM to reference specific chunks. Which two design choices are required to produce reliable citations? (Choose two.)

Choose an answer to begin — your selection is scored in the full session.

Accepted Answer

Store a stable document identifier and chunk metadata alongside each embedding in the Vector Search index.. Reliable citations require two things: the retriever must return source identifiers, and those identifiers must reach the LLM in the prompt. Storing document IDs and chunk metadata in the Vector Search index enables the first, and including them in the prompt context enables the second. Similarity metrics, temperature, and inference tables do not provide the linkage needed for citations.

Answer

Increase the LLM temperature to encourage the model to generate more detailed citations.

Answer

Enable inference tables on the Vector Search endpoint to capture citation metadata.

Answer

Set the Vector Search index to use a cosine similarity metric instead of dot product.

Question 2

A developer is building an agentic workflow using Databricks and LangChain. The agent needs to decide whether to answer a user's query directly or call an external tool to retrieve additional information. The developer wants to ensure the agent's decisions are logged for debugging and auditing. Which approach should they take to achieve this?

Accepted Answer

Use MLflow's autologging for LangChain to automatically capture agent steps and tool calls.. MLflow autologging for LangChain is the most efficient and integrated way to capture agent steps. It automatically logs tool calls, chain sequences, and inputs/outputs as MLflow traces, which can be inspected in the MLflow UI. This provides auditability and simplifies debugging without manual instrumentation.

Answer

Use Databricks Jobs to schedule the agent and rely on job run history for debugging.

Answer

Manually log each tool call using print statements to the driver logs.

Answer

Implement a custom logging function that writes agent decisions to a Delta table.

Question 3

A GenAI engineer has built a RAG application that queries a Delta table named `prod_ai.knowledge_base.documents` through a Databricks SQL warehouse. The application uses a service principal `sp-rag-prod` to authenticate. The table contains confidential internal documents. The security team wants to ensure that the service principal can only read the table and cannot modify or delete it. Which Unity Catalog privilege should be granted to `sp-rag-prod` on the table?

Accepted Answer

SELECT. The service principal needs only to read the table for the RAG application. Granting SELECT provides exactly that capability without allowing data modification. Other privileges either grant excessive rights (MODIFY, ALL PRIVILEGES) or do not permit reading table data (USE SCHEMA). SELECT is the correct least-privilege grant for a read-only consumer in Unity Catalog.

Answer

ALL PRIVILEGES

Answer

USE SCHEMA

Answer

MODIFY

Question 4

A developer is packaging a GenAI chat application as an MLflow model that will be deployed to a Mosaic AI Model Serving endpoint. The application needs to load a retrieval index and a prompt template at startup so the first request is not slowed by initialization. Which MLflow logging pattern should the developer use?

Accepted Answer

Log the application with mlflow.pyfunc.log_model and perform the heavy initialization in the model's load_context method so resources are ready before predict is called.. MLflow's PyFunc flavor calls load_context when the model is loaded, which happens before the endpoint begins serving requests. Initializing the retrieval index and prompt template there ensures they are ready once per replica and reused across calls. Loading inside predict or fetching resources per request would add latency, and a signature alone cannot prepare application state.

Answer

Use mlflow.pyfunc.log_model with a signature only, and rely on the serving environment's default warm-up to populate the index and template automatically.

Answer

Log only the prompt template as an artifact and let the Model Serving endpoint fetch the index from Unity Catalog on each request.

Answer

Log the application with mlflow.pyfunc.log_model and load the index and template inside the predict function on every request.

Question 5

A team wants its Databricks App to read a secret that stores an external API key used by an agent tool. The secret is managed in a Databricks secret scope. Which approach correctly exposes the secret to the running app without putting the value in source control?

Accepted Answer

Reference the secret scope and key in the app configuration's environment variables using the secrets-backed value syntax, then read it from the environment in code.. The app configuration can declare environment variables whose values are resolved from a Databricks secret scope using the secrets-backed reference syntax. The platform injects the resolved value at runtime, so the key is available to the process without ever being written into source, committed to a repository, or surfaced through deployment output.

Answer

Pass the key as a bundle variable in databricks.yml so it is substituted during deployment.

Answer

Commit the API key into a .env file in the app repository and load it with a dotenv library at startup.

Answer

Store the key in a Delta table and have the app query the table with its service principal at startup.

Question 6

An AI engineer is using MLflow to track experiments for a generative AI application. They want to log parameters, metrics, and artifacts for each run, and later compare runs to select the best model. Which MLflow component should they use to organize runs into a named group for a specific project?

Accepted Answer

MLflow Experiment. MLflow Experiments are designed to group runs for a specific project. They allow you to log and compare multiple runs, making it easy to track progress and select the best model. The experiment is the primary organizational unit in MLflow, and it is where runs are created and stored.

Answer

MLflow Model Registry

Answer

MLflow Run

Answer

MLflow Tracking Server

Question 7

A team is developing a generative AI application that uses an LLM to answer questions based on internal documents. They want to ensure the application is robust and provides accurate responses. Which TWO practices should they implement to improve the reliability of the application? (Choose two.)

Accepted Answer

Implement retrieval-augmented generation (RAG) to ground the LLM's responses in the internal documents.. The two practices are RAG and guardrails. RAG grounds responses in internal documents, reducing hallucinations and providing accurate, up-to-date information. Guardrails validate outputs to filter out harmful or incorrect content. Together, they significantly improve the reliability and safety of a GenAI application, especially for enterprise use cases.

Answer

Fine-tune the LLM on the internal documents to embed knowledge directly into the model weights.

Answer

Use a larger LLM with more parameters to increase the likelihood of correct answers.

Answer

Set the LLM's temperature to a high value to encourage more creative and diverse responses.

Question 8

A developer is creating a Databricks notebook to orchestrate a GenAI pipeline that includes data ingestion, vector index refresh, and model inference. They want to ensure that the pipeline can be easily tested and deployed across different environments. Which Databricks feature should they use to define the pipeline as code and manage deployments?

Accepted Answer

Databricks Asset Bundles (DABs).. Databricks Asset Bundles provide a declarative way to define and deploy Databricks resources, including jobs, notebooks, and configurations, as code. They support environment-specific parameters and CI/CD integration, making them the best choice for managing a GenAI pipeline across multiple environments.

Answer

Databricks Jobs with notebook tasks and parameters.

Answer

Delta Live Tables with expectations.

Answer

MLflow Projects with a conda environment.

Question 9

A team maintains a Mosaic AI Agent application whose endpoint must call an external LLM provider through a secret stored in a Databricks secret scope. During a deployment pipeline run, the endpoint build step fails while resolving the credential, even though the secret scope exists and the notebook test works. Which configuration should the engineer verify first?

Accepted Answer

That the serving endpoint's service principal or the deploying user has READ permission on the secret scope and that the code reads the secret at load time using the correct scope and key.. Credential resolution in Model Serving depends on the identity that loads the model having READ access to the secret scope and on the code requesting the correct scope and key. A notebook author with access may succeed while the endpoint identity fails, and a typo in scope or key produces the same symptom. Checking permissions and the retrieval call resolves the failure without restructuring the deployment.

Answer

That the secret scope is replicated into the Unity Catalog metastore attached to the serving endpoint.

Answer

That the endpoint has been configured with a personal access token embedded directly in the model signature.

Answer

That the secret scope is mounted as a Databricks secret scope in the serving endpoint's environment variables.

Question 10

A team deploys a Mosaic AI Agent application through Databricks Apps and must let the app call a Model Serving endpoint that enforces Unity Catalog permissions on the underlying model. Which TWO configurations are required for the app to authenticate and be authorized to query that endpoint? (Choose two.)

Accepted Answer

Grant the app's service principal the appropriate privileges on the serving endpoint and the model it serves.. Querying a Unity Catalog-governed serving endpoint requires both a valid identity and the right privileges. The app must present an OAuth token for its service principal so the request is authenticated, and that service principal must be granted access to the endpoint and the served model so authorization succeeds. Together these let the app call the endpoint without embedding human credentials.

Answer

Register the app as a Delta Sharing recipient and share the model with it.

Answer

Enable public network access on the serving endpoint so the app can reach it without a token.

Answer

Embed a personal access token for a workspace admin in the app source so the endpoint always resolves an admin identity.

Question 11

A generative AI engineer is packaging a retrieval-augmented generation (RAG) application so it can be deployed as a Databricks App. The app reads the vector index name and the serving endpoint name from environment configuration so the same code can run in dev and prod. Which approach correctly supplies these values at deploy time using Databricks Asset Bundles?

Accepted Answer

Declare bundle variables with defaults in databricks.yml, override them per target, and reference them in the app resource configuration.. Bundle variables with defaults in databricks.yml, overridden per target and referenced in the app resource, let one codebase deploy to dev and prod with different endpoint and index names. The bundle resolves values at deploy time and validates them, so promotion requires no source edits and each target receives the correct configuration without runtime prompts or extra lookups.

Answer

Store the values in a notebook widget and have the app prompt the user for them at startup.

Answer

Hard-code the dev and prod values in the app source and select a branch during deployment.

Answer

Write the values into a Delta table and have the app query that table on every request.

Question 12

A GenAI engineer is building a Databricks RAG application that answers questions over a Delta table containing 40 million support tickets. Users report that simple keyword lookups return irrelevant results because the tickets use inconsistent terminology. The engineer needs semantic retrieval that stays synchronized as new tickets stream in every few minutes. Which Databricks component should be used to serve this retrieval layer?

Accepted Answer

A Databricks Vector Search index with a Delta Sync index over the tickets table. Semantic retrieval over frequently changing data requires both embedding-based similarity and automatic synchronization with the source table. A Databricks Vector Search Delta Sync index provides exactly this: it computes and maintains embeddings as the Delta table changes and exposes a low-latency endpoint for similarity queries, so terminology variations are handled by vector proximity rather than literal string matching.

Answer

A Feature Store table that materializes ticket embeddings refreshed by a nightly job

Answer

A Delta Live Tables pipeline that computes embeddings and writes them to a Parquet directory

Answer

A Databricks SQL warehouse running a LIKE query over the ticket text column

Question 13

An engineer is deploying a RAG application whose retrieval step calls an external vector database over the public internet. The serving endpoint must reach that database, but corporate policy forbids exposing credentials in the model artifact and forbids public egress from the serving environment. Which configuration satisfies both constraints?

Accepted Answer

Store the database credential in a Databricks secret scope, reference it from the endpoint configuration, and enable network connectivity through a private link or VPC configuration for the serving endpoint.. Credentials belong in a Databricks secret scope so they are injected at runtime rather than stored in the artifact, and reaching a private external service from Model Serving requires explicit private connectivity rather than default public routing. Satisfying both constraints means combining secret injection with a private network path.

Answer

Have the retrieval function read the credential from a Delta table at query time and rely on the default public internet route for database access.

Answer

Embed the database credential as an environment variable inside the logged MLflow model so the serving container can read it at startup.

Answer

Package the credential into a Python wheel installed as a model dependency and configure the endpoint with serverless compute only.

Question 14

A team is designing a GenAI application that must call an external LLM provider from a Databricks notebook and from a Model Serving endpoint. The security team requires that the provider API key never appear in notebook source code, Git history, or the model artifact, and that the same governed credential be reused by both the notebook and the serving endpoint. Which design satisfies these requirements?

Accepted Answer

Store the API key in a Unity Catalog connection that supports external services, and reference that connection from both the notebook and the Model Serving endpoint.. Centralizing the external provider credential in a Unity Catalog connection keeps plaintext out of source code, Git history, and artifacts while letting both notebooks and Model Serving endpoints reference the same governed object. Secret scopes with manual injection, widget defaults, and cluster Spark configurations either leak the value or fail to reach the serving runtime.

Answer

Hard-code the API key as a widget default in the notebook and mirror it into the serving endpoint's environment variables so both components share one value.

Answer

Store the API key in a Databricks secret scope and reference it with dbutils.secrets.get in the notebook, then pass the resolved value into the serving endpoint's environment variables.

Answer

Place the API key in a cluster-scoped Spark configuration on the interactive cluster, and have the serving endpoint read the same configuration at request time.

Question 15

A team is preparing a Databricks Asset Bundle that deploys a GenAI application consisting of a job that builds a vector index and a Model Serving endpoint that hosts the agent. Before merging, they want the pipeline to validate and deploy the bundle to a staging workspace automatically. Which TWO bundle capabilities should the pipeline rely on? (Choose two.)

Accepted Answer

Run databricks bundle deploy with the staging target to create or update the job, endpoint, and related resources in the staging workspace.. A bundle pipeline validates the configuration against the target workspace and then deploys it to that target. Validation catches malformed or unresolvable configuration before anything changes, and deployment applies the resolved resources to the staging workspace, which together form the automated promotion path the team wants.

Answer

Run databricks bundle schema to regenerate the JSON schema and commit it to the repository.

Answer

Run databricks bundle run to execute the vector index job as part of the merge validation.

Answer

Run databricks bundle generate to create resource definitions from existing workspace objects before each deployment.