Databricks GenAI Assoc study guide · 2026

The 6 Databricks GenAI Assoc domains, explained

6 domains and 56 official objectives. Application Development carries the most weight at 30%. Each guide explains what the domain covers, how its questions are framed and where candidates lose marks.

Start with Domain 3 See all domains
Questions
45 scored (plus up to 10 unscored)
Time
90 minutes
Passing score
Not published by Databricks
Outline
Exam guide for the version live 18 March 2026 (current; guide PDF reissued August 2026)

Exam blueprint

Bar widths follow the official weights. Weight ranges are drawn at their midpoint.

  1. 1 14%
  2. 2 14%
  3. 3 30%
  4. 4 22%
  5. 5 8%
  6. 6 12%
  1. 1 Design Applications
  2. 2 Data Preparation
  3. 3 Application Development
  4. 4 Assembling and Deploying Applications
  5. 5 Governance
  6. 6 Evaluation and Monitoring

All 6 domains

56 official objectives 50–69 hrs suggested study

  1. Domain 01 14%

    Design Applications

    This section covers the design decisions made before any code is written: shaping prompts for structured output, mapping business needs to model tasks and chain components, ordering agent tools, and knowing when an Agent Bricks template already solves the problem.

    Inside this domain

    • Design a prompt that elicits a specifically formatted response
    • Select model tasks to accomplish a given business requirement
    • Select chain components for a desired model input and output
    • Translate business use case goals into a description of the desired inputs and outputs for the AI pipeline
    • and 2 more objectives

    Watch for Persona fixes format

    Objectives
    6
    Questions est.
    ~6
    Study time est.
    6–9 hrs
  2. Domain 02 14%

    Data Preparation

    This section covers getting documents ready for retrieval: choosing sources, parsing and cleaning content, chunking to fit model limits, writing chunks to Unity Catalog Delta tables, and measuring and improving what the retriever returns.

    Inside this domain

    • Apply a chunking strategy for a given document structure and model constraints
    • Filter extraneous content in source documents that degrades quality of a RAG application
    • Choose the appropriate Python package to extract document content from provided source data and format
    • Define operations and sequence to write given chunked text into Delta Lake tables in Unity Catalog
    • and 4 more objectives

    Watch for Bigger top-k fixes noise

    Objectives
    8
    Questions est.
    ~6
    Study time est.
    7–10 hrs
  3. Domain 03 30%

    Application Development

    The largest section covers building the application: frameworks, prompt augmentation and adjustment, guardrails, choosing LLMs and embedding models, reading model cards and experiment metrics, MLflow with Agent Framework, and connecting agents to Genie spaces.

    Inside this domain

    • Select Langchain/similar tools for use in a Generative AI application
    • Qualitatively assess responses to identify common issues such as quality and safety
    • Select chunking strategy based on model & retrieval evaluation
    • Augment a prompt with additional context from a user's input based on key fields, terms, and intents
    • and 9 more objectives

    Watch for Biggest model wins

    Objectives
    13
    Questions est.
    ~13–14
    Study time est.
    14–18 hrs
  4. Domain 04 22%

    Assembling and Deploying Applications

    This section covers turning a working prototype into a served application: pyfunc chains, Unity Catalog registration, Vector Search indexes, Foundation Model APIs, batch inference with ai_query, memory stores, CI/CD, MCP servers, prompt versioning and user interfaces.

    Inside this domain

    • Code a chain using a pyfunc model with pre- and post-processing
    • Control access to resources from model serving endpoints
    • Code a simple chain according to requirements
    • Choose the basic elements needed to create a RAG application: model flavor, embedding model, retriever, dependencies, input examples, model signature
    • and 11 more objectives

    Watch for Stages over aliases

    Objectives
    15
    Questions est.
    ~10
    Study time est.
    12–16 hrs
  5. Domain 05 8%

    Governance

    The smallest section covers keeping a generative AI application safe and lawful: masking sensitive data, guarding against malicious prompts, respecting data source licenses, and replacing or cleaning problematic text in the data that feeds the application.

    Inside this domain

    • Use masking techniques as guard rails to meet a performance objective
    • Select guardrail techniques to protect against malicious user inputs to a Gen AI application
    • Use legal/licensing requirements for data sources to avoid legal risk
    • Recommend an alternative for problematic text mitigation in a data source feeding a GenAI application

    Watch for Delete instead of mask

    Objectives
    4
    Questions est.
    ~3–4
    Study time est.
    4–6 hrs
  6. Domain 06 12%

    Evaluation and Monitoring

    This section covers proving and keeping quality: choosing models from evaluation metrics, MLflow scoring and tracing, judges that need ground truth, custom scorers, inference tables, AI Gateway usage tracking, cost control and expert feedback loops.

    Inside this domain

    • Select an LLM choice (size and architecture) based on a set of quantitative evaluation metrics
    • Select key metrics to monitor for a specific LLM deployment scenario
    • Evaluate agent performance using MLflow scoring and tracing
    • Use inference logging to assess deployed RAG application performance
    • and 6 more objectives

    Watch for Correctness without labels

    Objectives
    10
    Questions est.
    ~5–6
    Study time est.
    7–10 hrs

Study the heaviest domains first

Marks follow weight. This order puts the most marks behind you first, across about 50–69 hrs of study.

  1. 3 Application Development 30% · 14–18 hrs
  2. 4 Assembling and Deploying Applications 22% · 12–16 hrs
  3. 1 Design Applications 14% · 6–9 hrs
  4. 2 Data Preparation 14% · 7–10 hrs
  5. 6 Evaluation and Monitoring 12% · 7–10 hrs
  6. 5 Governance 8% · 4–6 hrs

Try one

Practice questionDomain 3: Application Development

A support agent built with LangGraph sometimes gives wrong refund amounts. The team cannot tell whether the error comes from the order lookup tool, the retriever or the final LLM call. What should they use to find the faulty step?

One trap from every domain

The answer that looks right, and the clue that gives it away.

  • Domain 1

    Persona fixes format

    Clue The problem described is unparseable output, not tone.

  • Domain 2

    Bigger top-k fixes noise

    Clue Retrieved text is full of repeated headers or menus.

  • Domain 3

    Biggest model wins

    Clue The scenario stresses latency, throughput or budget.

  • Domain 4

    Stages over aliases

    Clue The model is registered in Unity Catalog.

  • Domain 5

    Delete instead of mask

    Clue The requirement mentions keeping answer quality or coverage.

  • Domain 6

    Correctness without labels

    Clue The data has no reference answers.

Other Artificial Intelligence & Machine Learning certifications

Study guides for exams that overlap with Databricks GenAI Assoc.

Practise what you read

Databricks GenAI Assoc domains: common questions

How many domains are in the Databricks GenAI Assoc exam?

Databricks GenAI Assoc has 6 domains with 56 official objectives between them: Design Applications, Data Preparation, Application Development, Assembling and Deploying Applications, Governance and Evaluation and Monitoring.

Which Databricks GenAI Assoc domain carries the most weight?

Application Development is the heaviest at 30%, followed by Assembling and Deploying Applications (22%). The lightest is Governance (8%). Weights come from Databricks's official outline.

What is the Databricks GenAI Assoc exam format?

Questions: 45 scored (plus up to 10 unscored). Time allowed: 90 minutes. Passing score: Not published by Databricks.

How long should I study for Databricks GenAI Assoc?

Our planning estimate is 50–69 hrs across all domains, split by weight. It is an editorial estimate, not an official figure, so adjust it to your experience and to how you score on practice questions.

Where should I start?

Start with Application Development, because it carries the most marks, then work down by weight. Use the checklist on each domain page to confirm you can do what the domain asks before moving on.

Which outline is this guide based on?

This guide follows Exam guide for the version live 18 March 2026 (current; guide PDF reissued August 2026), published by Databricks. It was last reviewed in October 2026.

Domain names, weights and objectives come from Databricks's official outline (Exam guide for the version live 18 March 2026 (current; guide PDF reissued August 2026)). Question counts and study hours are editorial planning estimates, not official figures. Last reviewed October 2026.