Databricks Certified Generative AI Engineer Associate Practice Exam

What the Databricks Certified Generative AI Engineer Associate exam covers

• Design Applications96 questions
• Data Preparation95 questions
• Application Development176 questions
• Assembling and Deploying Applications152 questions
• Governance84 questions
• Evaluation and Monitoring108 questions

Free Databricks Certified Generative AI Engineer Associate sample questions

A sample of 10 questions with answers and explanations. Sign up free to practice all 711.

1. Question 1Application Development

   A retail support bot uses Mosaic AI Vector Search to fetch product return policies. Before sending the user's question to the LLM, the app must combine the retrieved passages with the original question so the model answers using only that context. Which prompt construction pattern accomplishes this?

   • AInsert the retrieved passages into the prompt as a context block alongside the questionCorrect
   • BReference the retrieved passages by their Unity Catalog volume path in the question text
   • CReplace the user's question with only the single top retrieved passage
   • DSend the user question alone and rely on the LLM's pretrained knowledge of the policy
   ✓ Correct answer: A

   Vector Search returns passages, but the LLM only sees what is actually included in the prompt tokens sent to it. The standard pattern concatenates the retrieved chunks with the user's question, usually under a labeled context section, so the model can cite and reason over that material.

   Why the other options are wrong

   • BA file path is not readable content; the model cannot open a volume reference on its own.
   • CDiscarding the original question loses the user's actual intent, breaking the interaction.
   • DRelying on pretrained knowledge skips retrieval entirely and risks outdated or wrong answers.
2. Question 2Application DevelopmentSelect all that apply

   An engineer selecting a model from a hub for a healthcare-adjacent application notices the model card includes a disclaimer that the model is not validated for clinical or diagnostic use. Which TWO actions are appropriate given this disclaimer? (Choose TWO)

   • AIgnore the disclaimer, since model cards are only informal suggestions
   • BAssume the disclaimer only applies to competing organizations, not this project
   • CFine-tune the model once to permanently remove the disclaimer's restriction
   • DAvoid using the model for diagnostic decision-making, consistent with the documented limitationCorrect
   • ETreat the disclaimer as a legitimate constraint when scoping the application's intended use caseCorrect
   ✓ Correct answer: D, E

   The appropriate response to a disclosed clinical-use limitation is to keep the application within the model's validated scope and to treat that boundary as a real constraint on the project, not an optional note.

   Why the other options are wrong

   • AIgnoring a documented limitation is an unsound and risky practice.
   • BDisclaimers apply based on the stated use case, not on which organization is using the model.
   • CFine-tuning does not remove a vendor's documented limitation or associated liability disclaimer.
3. Question 3Application Development

   A team notices their agent's quality metrics were strong at launch but wants ongoing visibility into whether real user traffic is still getting good answers weeks later. Which phase of the GenAI life cycle addresses this ongoing need?

   • AMonitoringCorrect
   • BEvaluation
   • CData preparation
   • DApplication design
   ✓ Correct answer: A

   Unlike evaluation, which is a point-in-time, pre-deployment comparison, monitoring runs continuously against live inference data, often via inference tables and Lakehouse Monitoring, to answer whether the system is still behaving well now. Data preparation and application design are earlier life cycle phases concerned with building the data pipeline and initial architecture, not with ongoing production observability.

   Why the other options are wrong

   • BEvaluation is a point-in-time, typically pre-deployment comparison of versions, not an ongoing production observability practice.
   • CData preparation concerns building and cleaning the underlying data pipeline earlier in the life cycle, not production monitoring.
   • DApplication design is the early architectural phase of the life cycle, unrelated to ongoing observation of live traffic.
4. Question 4Data Preparation

   A data engineer installs the pytesseract Python package but OCR calls fail with an error that the OCR engine cannot be found. What is the most likely cause?

   • AThe Delta table schema does not allow string columns, unrelated to a missing OCR engine binary
   • BThe Vector Search index has not been created yet
   • Cpytesseract wraps the Tesseract OCR engine, which must be separately installed on the systemCorrect
   • DThe Unity Catalog volume is set to read-only
   ✓ Correct answer: C

   Installing the pytesseract Python package alone does not install the underlying Tesseract executable it calls out to. On many systems this binary must be installed separately (for example through the OS package manager) and be discoverable on the system path.

   Why the other options are wrong

   • ADelta table schema constraints have no connection to an OCR engine lookup error.
   • BA missing Vector Search index would surface as a different error at query time, not an OCR engine error.
   • DA read-only volume would cause a write failure, not an engine-not-found error.
5. Question 5Assembling and Deploying Applications

   A team needs a simple chain that answers customer questions using only the current product catalog. Which sequence correctly assembles the chain?

   • ASend the question directly to the LLM, then retrieve catalog chunks afterward to append as a footnote
   • BRetrieve catalog chunks, send them straight to the LLM as the entire prompt, and ignore the original question
   • CInsert the question into the prompt template, call the LLM, and only retrieve catalog chunks if the answer looks wrong
   • DRetrieve the most relevant catalog chunks, insert them into the prompt template with the question, then call the LLMCorrect
   ✓ Correct answer: D

   Retrieval must happen before the prompt is built so the retrieved chunks can be substituted into the template alongside the user's question; the LLM is called last, on the fully assembled prompt.

   Why the other options are wrong

   • ACalling the LLM before retrieval means the answer is generated with no grounding, and appending chunks afterward does not let the model use them.
   • BSending only chunks with no question gives the LLM no task to perform against that context.
   • CCalling the LLM before retrieval, with retrieval only as an afterthought, defeats the purpose of grounding answers in current data.
6. Question 6Assembling and Deploying Applications

   A team receives embedding vectors in real time from an external streaming system that does not write to a Delta table, and needs to upsert and delete individual vectors on demand via API. Which Vector Search index type fits this requirement?

   • AA Delta Sync index
   • BA Model Serving batch endpoint
   • CA Genie space
   • DA direct-access indexCorrect
   ✓ Correct answer: D

   Unlike a Delta Sync index, which requires and automatically syncs from a source Delta table, a direct-access index has no such backing table and is written to directly through the Vector Search API. This matches a scenario where vectors arrive from an external streaming system.

   Why the other options are wrong

   • AA Delta Sync index requires a backing Delta table it automatically syncs from, which does not exist here.
   • BA batch serving endpoint is for scoring workloads, not for storing or querying vectors.
   • CA Genie space is a natural-language SQL interface, unrelated to vector storage.
7. Question 7Design Applications

   Output from an LLM call needs to be loaded into a Python dict by downstream code with no manual cleanup. Which prompt design habit best avoids parsing failures?

   • AAsk for raw JSON only, with no markdown code fences or surrounding explanation, matching a stated schemaCorrect
   • BAsk the model to wrap the JSON in a markdown code block so it is easy for a person to read
   • CAsk the model to append a short natural-language summary after the JSON for extra context
   • DAsk the model to bold the key names so the structure is visually clear in the response
   ✓ Correct answer: A

   Markdown fences, bold markup, or trailing commentary all introduce characters that break a strict JSON parser unless extra cleanup code is added; asking for raw JSON matching a schema removes that failure mode at the source.

   Why the other options are wrong

   • BMarkdown code fences add characters around the JSON that a strict parser will fail on unless stripped first.
   • CTrailing prose after the JSON object breaks naive parsers that expect the JSON to be the entire response.
   • DBold markup injects non-JSON characters into what should be a clean, machine-readable payload.
8. Question 8Design Applications

   A content-moderation step must check a generated response for safety issues using a guardrail tool before it reaches the user. What is the correct placement of this check in the pipeline?

   • ARun the safety check before any response has been generated, so nothing exists yet to check
   • BDeliver the response to the user first, then run the safety check afterward as a formality
   • CGenerate the response and deliver it immediately, skipping any safety check
   • DGenerate the response, run a safety check on it, then deliver it only if the check passesCorrect
   ✓ Correct answer: D

   The guardrail tool's input is the generated response itself, so generation must happen first; the check must then run and gate delivery, since checking after the user already received the response defeats the purpose of the safeguard.

   Why the other options are wrong

   • AThere is nothing to check before a response has actually been generated.
   • BChecking after the user already saw the response cannot prevent an unsafe response from being seen.
   • CSkipping the safety check entirely removes the safeguard the design explicitly calls for.
9. Question 9Evaluation and Monitoring

   A team evaluated a 70B model and a 7B model and found the 7B model meets their quality bar for a simple intent-classification task at a fraction of the cost. What Databricks cost-control practice does choosing the 7B model represent?

   • AProvisioned throughput reservation
   • BAI Gateway rate limiting
   • CResponse caching
   • DRight-sizing the model to the taskCorrect
   ✓ Correct answer: D

   Right-sizing means selecting the smallest/cheapest model that still meets the measured quality bar for a given task, which is exactly what choosing the 7B model over the 70B model for adequate intent-classification accuracy represents.

   Why the other options are wrong

   • AThis scenario is about model choice, not a capacity reservation mechanism.
   • BRate limiting controls request volume, not model selection.
   • CCaching avoids redundant calls, unrelated to choosing a smaller architecture.
10. Question 10Governance

    A telecom customer-care bot must mask phone numbers in retrieved call-log excerpts, but the masking should vary: a support rep sees the last four digits, a compliance auditor sees the full number, and a customer talking to the bot sees nothing. What mechanism natively supports this three-tier disclosure?

    • AThree separate physical copies of the call-log table
    • BA single AI Gateway rate limit rule configured with three tiers
    • CA column mask function with conditional logic keyed to group membershipCorrect
    • DThree Vector Search indexes built with different embedding models
    ✓ Correct answer: C

    Unity Catalog column masks support conditional expressions, so the same underlying phone-number column can return full, partial, or fully redacted values depending on who is querying, without maintaining duplicate copies of the data.

    Why the other options are wrong

    • APhysical duplication is wasteful and hard to keep in sync as data changes.
    • BRate limiting controls request volume, not data visibility per role.
    • DMultiple indexes with different embedding models addresses retrieval quality, not tiered masking.

Who this Databricks Certified Generative AI Engineer Associate practice exam is for

This practice set is for anyone preparing for the Databricks Certified Generative AI Engineer Associate exam at the intermediate level - from first-time candidates building a foundation to experienced Data practitioners doing a final review before test day. If you learn best by working through realistic questions and reading why each answer is right or wrong, it is built for you.

How to use this Databricks Certified Generative AI Engineer Associate practice exam

1. Start with the free sample questions above to gauge your current baseline.
2. Read the full explanation on every question, including why each wrong option is wrong.
3. Track your weak domains and focus your study where you are losing the most marks.
4. Once you are scoring consistently well, take a timed, full-length mock exam.
5. Use your readiness score to decide when you are ready to book the real Databricks Certified Generative AI Engineer Associate exam.

Related Data resources

• Databricks Certified Generative AI Engineer Associate study guideKey concepts
• Data practice examsAll Data
• Certification pathWhere this fits
• Certification exam guides & tipsBlog
• Plans & pricingFree & paid
• Databricks Certified Data Analyst Associate practice examRelated
• MongoDB Associate Developer practice examRelated
• Databricks Certified Data Engineer Associate practice examRelated

Databricks Certified Generative AI Engineer Associate practice exam FAQ