Databricks Certified Generative AI Engineer Associate Study Plan

A practical 7-day, 14-day, 30-day, and 60/90-day study plan for the Databricks Certified Generative AI Engineer Associate exam.

Orientation

This Study Plan is for candidates preparing for the Databricks Certified Generative AI Engineer Associate exam, official exam code GenAI Engineer, from Databricks.

Use it to turn your remaining time into a practical schedule for generative AI engineering topics on Databricks: retrieval-augmented generation, vector search, embeddings, prompt design, model serving, evaluation, governance, security, MLflow/Mosaic AI workflows, troubleshooting, and production-readiness decisions.

This is an independent study planning resource. Always use the current Databricks exam guide as your final scope reference.

Which plan should you use?

Time availableBest forPrimary goalMock exam use
7 daysYou have studied before or work with Databricks regularlyFinal review, weak-area repair, timed practice1 timed mock early, 1 final timed set
14 daysYou know GenAI concepts but need Databricks-specific reviewFocused domain coverage plus hands-on reinforcementDiagnostic on Day 1, timed mock near Days 10-12
30 daysYou want a balanced plan while working full timeBuild coverage, practice scenarios, review missesDiagnostic Week 1, mocks in Weeks 3 and 4
60/90 daysYou are newer to Databricks GenAI engineeringFull preparation with labs, notes, and repeated practiceMonthly checkpoint mocks, final-week timed mock

Core exam-prep priorities

Prioritize the work that most directly supports exam decisions.

Priority areaWhat to be able to do
Generative AI architectureChoose patterns for RAG, tool use, agents, chat applications, and evaluation loops
Databricks platform workflowUnderstand how notebooks, jobs, model serving, vector search, governance, and monitoring fit together
Retrieval-augmented generationExplain chunking, embeddings, vector indexes, retrieval quality, grounding, and hallucination reduction
Model serving and endpointsKnow when to use hosted model serving, foundation model APIs, endpoint configuration concepts, and operational tradeoffs
EvaluationCompare expected output quality, retrieval quality, safety, latency, cost, and regression checks
MLflow and Mosaic AI conceptsTrack prompts, parameters, traces, evaluations, model versions, and application behavior where applicable
Security and governanceApply Unity Catalog concepts, data permissions, access controls, secret handling, and safe data use
TroubleshootingDiagnose poor responses, irrelevant retrieval, latency, permissions, deployment errors, and evaluation failures

Daily practice rhythm

Use this rhythm on most study days. Adjust the duration, not the order.

Block45-minute day90-minute day2-3 hour day
Warm-up recall5 min10 min15 min
Learn or review one topic15 min25 min40 min
Hands-on or scenario practice15 min30 min45-60 min
Practice questions5-7 min15 min25-35 min
Missed-question review5 min10 min20 min
Update weak-area list2 min5 min5-10 min

Daily rule: every study session should produce one of these outputs:

  • A corrected misunderstanding.
  • A short architecture decision note.
  • A reviewed set of missed questions.
  • A hands-on note explaining what failed and how you fixed it.
  • A flashcard or checklist item for final review.

Start with a diagnostic

Before choosing what to study, find your gaps.

Diagnostic setup

StepAction
1Take a mixed practice set without notes.
2Mark each question as confident, guessed, or unfamiliar.
3Review every missed and guessed question.
4Tag each miss by topic: RAG, embeddings, vector search, serving, evaluation, MLflow, security, governance, troubleshooting, or architecture.
5Build a weak-area list and use it to drive the next 3-5 study sessions.

Diagnostic scoring categories

Do not only track right and wrong answers. Track why you missed.

Miss typeMeaningFix
Concept gapYou did not know the ideaStudy the topic from the exam guide and create a short note
Databricks workflow gapYou knew GenAI generally but not how it maps to DatabricksDo a focused platform review or mini-lab
Scenario misreadYou missed a key constraintPractice slower reading and underline requirements
Two-answer confusionYou narrowed it down but chose wrongWrite the decision rule that separates the answers
Terminology gapProduct or feature term was unclearBuild a term list and review it daily
OverengineeringYou picked a complex design when a simpler one fitPractice architecture tradeoff questions

7-day final review plan

Use this if the exam is within one week. Do not try to relearn everything. Your goal is to stabilize score, reduce careless misses, and repair the highest-value gaps.

DayMain focusStudy actionsOutput
1Diagnostic and triageTake a timed mixed set. Review every miss. Rank weak areas.Top 5 weak-area list
2RAG and retrievalReview embeddings, chunking, vector search, retrieval quality, grounding, and failure modes. Do targeted questions.RAG decision checklist
3Model serving and app flowReview model endpoints, foundation model usage patterns, latency/cost tradeoffs, deployment flow, and request path.Serving workflow note
4Evaluation and MLflowReview prompt evaluation, retrieval evaluation, regression testing, traces, logged parameters, and comparison of candidate approaches.Evaluation checklist
5Governance and securityReview Unity Catalog concepts, permissions, data access, secrets, PII/sensitive data handling, and safe GenAI application patterns.Security review sheet
6Timed mock and weak sprintTake a timed mock or large timed set. Review deeply. Rework only the weakest 2-3 topics.Final miss log
7Light final reviewReview notes, wrong-answer log, and decision rules. Do a small confidence set only. Stop heavy new study.Exam-day checklist

7-day rules

  • Stop adding new material after Day 5 unless it appears repeatedly in missed questions.
  • Do not spend Day 7 on a full new mock if it will leave you tired.
  • Revisit all guessed questions, even if they were correct.
  • Practice reading scenario constraints: data access, latency, quality, governance, deployment, and evaluation requirements.

14-day focused plan

Use this if you have two weeks and need both review and practice.

DayFocusActions
1DiagnosticMixed practice set, miss tagging, weak-area ranking
2Databricks GenAI workflowReview how data, prompts, models, endpoints, evaluation, and governance connect
3RAG foundationsChunking, embeddings, vector similarity, retrieval quality, hallucination reduction
4Databricks Vector Search and retrieval patternsReview index concepts, source data preparation, metadata filtering concepts, and troubleshooting poor retrieval
5Prompt engineeringInstructions, context, examples, output format, safety constraints, prompt iteration
6Model servingServing endpoint concepts, foundation model usage, latency, cost, scaling, deployment tradeoffs
7Review checkpointTimed practice set, review misses, update weak-area list
8EvaluationQuality metrics, human review, automated checks, regression testing, retrieval evaluation
9MLflow and trackingTrack prompts, parameters, versions, traces, evaluations, and compare experiments
10Security and governanceUnity Catalog concepts, access control, sensitive data, secrets, least privilege, auditability
11Architecture scenariosChoose designs for chat, RAG, summarization, classification, and knowledge assistant use cases
12Timed mockFull timed mock or largest available timed set; review all misses
13Weak-area sprintDrill the 2-4 topics with the most misses; retake similar questions
14Final reviewLight practice, review decision rules, prepare exam-day plan

14-day emphasis

Spend more time on scenario reasoning than memorization. The exam is likely to reward knowing which approach fits a given application requirement, not just recognizing product names.

30-day balanced plan

Use this if you can study consistently for a month. This is the best path for many working candidates.

Weekly structure

WeekGoalPractice target
Week 1Build baseline and cover platform workflowDiagnostic plus topic drills
Week 2Master RAG, embeddings, vector search, and prompt patternsScenario practice and mini-labs
Week 3Cover serving, evaluation, MLflow, governance, and troubleshootingTimed sets and hands-on review
Week 4Convert knowledge into exam readinessMock exams, weak-area sprint, final review

30-day schedule

DayFocusActions
1DiagnosticTake mixed set, tag misses, create study tracker
2Exam guide mappingMap official objectives to topics you know, partly know, and do not know
3Databricks GenAI workflowReview workspace flow, notebooks, data sources, endpoints, evaluation, governance
4Core GenAI conceptsLLM behavior, tokens conceptually, prompts, embeddings, grounding, hallucination risks
5RAG architectureRetrieval pipeline, chunking strategy, embedding selection concepts, answer generation
6Practice and reviewRAG and prompt questions; update miss log
7Weekly checkpointTimed set; review weak areas
8Vector searchIndex concepts, data preparation, update patterns, metadata and filtering concepts
9Retrieval troubleshootingIrrelevant results, stale data, poor chunking, missing permissions, weak prompts
10Prompt engineeringSystem/user instructions, examples, formatting, guardrails, evaluation prompts
11Application patternsChatbot, summarization, classification, Q&A, knowledge assistant scenarios
12Hands-on consolidationBuild or review a simple RAG flow conceptually in Databricks
13Practice and reviewTargeted questions on retrieval and architecture
14Weekly checkpointTimed set; update top weak areas
15Model servingEndpoint concepts, deployment flow, endpoint selection, operational considerations
16Foundation models and APIsWhen to use hosted foundation models, external models, or custom models
17MLflow and trackingExperiments, parameters, prompts, traces, model versions, comparisons
18EvaluationAnswer quality, retrieval quality, safety, human review, regression tests
19Governance and securityUnity Catalog, permissions, secrets, sensitive data, least privilege
20TroubleshootingLatency, permissions, retrieval failures, poor output, failed deployments
21Mock 1Timed mock or large timed set; deep review
22Mock reviewRework misses; write decision rules for repeated errors
23Weak area 1Target your largest miss category
24Weak area 2Target your second largest miss category
25Architecture scenariosPractice choosing between design options under constraints
26Security and evaluation reviewRevisit governance and evaluation because they affect many scenario questions
27Mock 2Timed mock or large timed set
28Final weak sprintDrill repeated misses; review guessed questions
29Final reviewReview notes, checklists, and decision rules; light timed set only
30Exam readinessRest, logistics, confidence set, no heavy new material

60/90-day full preparation path

Use this if you are newer to Databricks, newer to generative AI engineering, or want deeper hands-on practice.

60-day path

PhaseDaysGoalWhat to do
Foundation1-10Understand exam scope and GenAI basicsRead exam guide, take diagnostic, review LLMs, prompts, embeddings, RAG concepts
Databricks workflow11-20Connect GenAI concepts to DatabricksStudy notebooks, data preparation, model serving concepts, Unity Catalog, MLflow/Mosaic AI workflow
RAG depth21-32Build strong retrieval reasoningStudy chunking, embeddings, vector search, metadata, grounding, retrieval evaluation, failure modes
Serving and operations33-42Prepare for deployment decisionsStudy serving endpoints, foundation model usage, app patterns, latency/cost/governance tradeoffs
Evaluation and governance43-50Improve production-readiness judgmentReview evaluation, tracing, monitoring concepts, security, access control, safe data use
Exam conversion51-60Convert knowledge into timed performanceTimed mocks, weak-area sprints, final review, exam-day checklist

90-day path

Use the 60-day path, but add deeper practice between phases.

Added timeUse it for
Extra 10 days after FoundationHands-on notebooks, basic prompt experiments, terminology review
Extra 10 days after RAG depthBuild or review multiple RAG scenarios: internal docs, support Q&A, summarization with retrieval, metadata filters
Extra 10 days before Exam conversionMore timed mixed practice, governance scenarios, evaluation comparisons, troubleshooting drills

Long-path weekly cadence

Day typeActivity
2 days per weekLearn or review concepts
1-2 days per weekHands-on Databricks workflow or architecture walkthrough
1 day per weekPractice questions
1 day per weekMissed-question review and notes
Every 2-3 weeksTimed mixed checkpoint

Hands-on concept review checklist

You do not need to build a production system for exam prep, but you should understand how a GenAI application would be assembled and operated on Databricks.

AreaHands-on or walkthrough task
Data preparationIdentify source data, clean text, decide chunking approach, and explain metadata use
EmbeddingsExplain how documents and queries become vectors and why embedding choice matters
Vector searchWalk through index creation conceptually, retrieval, filters, freshness, and failure modes
RAG request pathTrace user question to retrieval to prompt construction to model response
Prompt iterationCompare a vague prompt with a constrained prompt that includes role, context, format, and refusal rules
Model servingExplain how an application calls a model endpoint and what operational constraints matter
EvaluationCompare two versions of a prompt or retriever and decide which is better
GovernanceIdentify who can access data, models, endpoints, and logs
TroubleshootingDiagnose poor answer quality, missing context, slow response, permission error, or unsafe output

RAG flow you should be able to explain

User question
  -> validate and prepare request
  -> embed query
  -> retrieve relevant chunks from vector index
  -> apply filters and ranking
  -> build prompt with instructions and context
  -> call model endpoint
  -> return answer with appropriate grounding or citations
  -> log trace, feedback, and evaluation signals

Use this flow for scenario questions. Ask: where is the failure happening, and which control fixes it?

Domain-by-domain study actions

Study taskQuestions to ask yourself
Review chunkingAre chunks too large, too small, or missing context?
Review embeddingsDoes the embedding model fit the data and query style?
Review metadataCan filters improve relevance or enforce scope?
Review groundingDoes the response use retrieved context or unsupported model knowledge?
Review freshnessHow are document updates reflected in retrieval?
Review troubleshootingIs the problem retrieval, prompt construction, permissions, or model behavior?

Prompt engineering

Study taskPractice action
Role and taskWrite the instruction so the model knows exactly what to do
Context boundariesSeparate retrieved context from user input
Output formatSpecify JSON, bullet list, short answer, or citation style when needed
Safety behaviorDefine what the model should do when context is insufficient
Few-shot examplesKnow when examples help and when they add noise
Prompt regressionCompare output before and after prompt changes

Model serving and application design

Scenario constraintDesign consideration
Low latencySimplify retrieval, reduce unnecessary calls, review endpoint and app design
Higher qualityImprove retrieval, prompts, evaluation, and feedback loop
Sensitive dataApply access control, governance, logging caution, and least privilege
Frequent updatesConsider data refresh and index update workflow
Multiple usersConsider endpoint access, scaling concepts, monitoring, and governance
Cost pressureAvoid unnecessary model calls, oversized context, and inefficient evaluation runs

Evaluation and MLflow/Mosaic AI review

TopicWhat to know
Offline evaluationCompare versions before release
Human evaluationUse reviewers for subjective quality or safety concerns
Retrieval evaluationCheck whether the right context is retrieved
Response evaluationCheck correctness, helpfulness, format, safety, and groundedness
Regression testingEnsure prompt or retriever changes do not break previous behavior
TrackingTrack prompts, parameters, versions, traces, metrics, and feedback where applicable

Security, governance, and responsible use

AreaReview points
Unity CatalogData governance, permissions, lineage concepts, managed access patterns
Access controlWho can read source data, query indexes, call endpoints, and view logs
SecretsAvoid hard-coded credentials and unsafe token handling
Sensitive dataConsider PII, proprietary data, and whether it should be used in prompts or logs
Least privilegeGrant only the access needed for the application or user
AuditabilityUnderstand why tracking, logs, and governance matter in GenAI systems
Safe responsesKnow how to handle insufficient context, unsafe requests, or unsupported claims

Missed-question review method

A missed-question log is more useful than rereading notes.

FieldWhat to record
DateWhen you missed it
TopicRAG, serving, evaluation, security, etc.
Question typeConcept, scenario, troubleshooting, terminology
Your answerWhat you chose
Correct ideaThe principle you should have applied
Why you missed itGap, misread, confusion, or overengineering
FixNote, flashcard, mini-lab, or retest
Retest dateWhen you will verify the fix

Review loop

  1. Review missed questions within 24 hours.
  2. Rewrite the correct decision rule in your own words.
  3. Add one similar practice question or scenario.
  4. Retest the topic 2-3 days later.
  5. Keep the item active until you can explain why the wrong answers are wrong.

Example decision-rule format

Weak areaDecision rule
Poor RAG answersFirst check retrieval quality and prompt grounding before blaming the model
Sensitive source dataApply governance and least privilege before designing the app flow
Irrelevant retrieved chunksRevisit chunking, embeddings, metadata filters, and index freshness
Unstable output after changesAdd evaluation and regression checks before release
Slow applicationCheck retrieval path, prompt size, model call pattern, and serving design

When to use timed mock exams

Timed mocks are for exam performance, not initial learning. Use them after you have reviewed enough content to learn from the results.

PlanFirst timed mockSecond timed mockFinal timed practice
7 daysDay 1 or 2Day 6Small confidence set Day 7
14 daysDay 7 or 8Day 12Light review Day 14
30 daysDay 21Day 27Small set Day 29
60/90 daysEvery 2-3 weeks after foundation phaseFinal 10 days2-3 days before exam

How to review a timed mock

Review stepAction
First passMark every miss and every guess
Second passCategorize by topic and miss type
Third passIdentify repeated patterns
Fourth passCreate 3-5 study tasks, not 20
Final passRe-answer missed questions without looking at the explanation

Avoid taking mock after mock without review. One deeply reviewed mock is usually more valuable than several unreviewed attempts.

Final-week rules

RuleWhy it matters
Stop adding broad new material 48-72 hours before the examPrevents overload and confusion
Keep studying weak areas, not favorite areasFavorite topics create false confidence
Review guessed correct answersThey reveal unstable knowledge
Use timed sets sparinglyProtect energy and focus
Practice scenario readingMany errors come from missing constraints
Sleep and logistics matterFatigue increases careless misses

Final 48-hour checklist

  • Review your top 10 decision rules.
  • Review all high-frequency missed topics.
  • Revisit RAG flow, evaluation flow, and governance flow.
  • Do one small timed set if it builds confidence.
  • Stop heavy study the evening before the exam.
  • Prepare identification, appointment details, workspace requirements, and timing plan.

Exam-readiness checks

You are likely ready when you can do the following without notes.

Readiness checkCan you do it?
Explain a Databricks GenAI application flow from source data to model responseYes / No
Diagnose poor RAG output using retrieval, prompt, data, and model factorsYes / No
Choose when vector search and embeddings are appropriateYes / No
Explain how evaluation supports prompt and application changesYes / No
Describe how MLflow/Mosaic AI concepts support tracking and comparisonYes / No
Apply governance and access-control thinking to GenAI scenariosYes / No
Handle timed questions without repeatedly running out of timeYes / No
Explain why the wrong answers are wrong on reviewed practice questionsYes / No

If several answers are “No,” do not simply read more. Pick the weakest area, do targeted practice, and review misses until the decision rule is clear.

Practical next step

Choose the plan that matches your exam date, take a diagnostic practice set, and build a missed-question log today. Use each study session to repair one specific weakness in your preparation for the Databricks Certified Generative AI Engineer Associate (GenAI Engineer) exam.