Amazon Q (Business and Developer)

Amazon Q is AWS's GenAI assistant family — Q Business (enterprise GenAI over company data, permission-aware) and Q Developer (AI coding assistant). AIF-C01 tests Q Business vs Kendra: Q Business returns synthesised answers; Kendra returns document passages.

Amazon Q was announced at re:Invent 2023 as AWS's answer to Microsoft Copilot and Google Duet. It comes in two distinct products that the AIF-C01 exam treats as separate services.


The Two Products

ProductWhat it doesWho uses it
Amazon Q BusinessGenAI assistant that answers questions using company dataEmployees, knowledge workers
Amazon Q DeveloperAI coding assistant for writing, debugging, and explaining codeDevelopers, DevOps

Amazon Q Business

A fully managed GenAI assistant that connects to enterprise data sources and answers natural language questions — with citations, permission enforcement, and audit trails.

How it works

  1. Connect data sources (S3, SharePoint, Confluence, Salesforce, ServiceNow, Jira, RDS, and 40+ connectors)
  2. Q Business indexes and chunks the content using Bedrock embeddings
  3. Users ask questions in natural language via the Q Business web UI or embedded widget
  4. Q Business retrieves relevant chunks, synthesises an answer using a foundation model, and provides source citations
  5. Answers are permission-aware: users only see content from sources they have access to

Key capabilities

  • Synthesised answers with citations — not just document links; generates a coherent response from multiple sources
  • Permission-aware retrieval — ACL enforcement means users cannot see content from systems they lack access to (enforced at query time, not just at index time)
  • Admin controls — topic blocklist, response guardrails, audit logging to CloudTrail
  • Plugins — take actions in third-party systems (create Jira tickets, update Salesforce records) from within Q Business
  • Customisation — custom system prompts, response styles, blocked topics

Amazon Q Business vs Amazon Kendra

This is the primary AIF-C01 exam trap:

DimensionAmazon Q BusinessAmazon Kendra
Output typeSynthesised GenAI answer with citationsRanked list of relevant document passages
TechnologyRAG over FM (Bedrock-powered)ML-powered semantic search (no FM generation)
Permission enforcementNative — ACL-aware retrievalNative — ACL-aware retrieval
Use case"What is our PTO policy?" → paragraph answer"Find documents about PTO policy" → document list
CustomisationSystem prompts, topic blocklist, pluginsCustom document attributes, facets, query tuning
Underlying modelAmazon Bedrock FM (selectable)ML ranking models (not generative)

Exam trigger for Q Business: "GenAI assistant over company data", "answer questions from SharePoint/Confluence", "synthesise answers from internal docs", "employee self-service AI"

Exam trigger for Kendra: "intelligent document search", "return relevant passages", "enterprise search across multiple repositories", "employees search company policies" (returns passages, does not synthesise)


Amazon Q Developer

An AI coding assistant integrated into IDEs and the AWS Console.

Capabilities

  • Code generation — generate code from natural language descriptions
  • Code completion — inline suggestions as you type (like GitHub Copilot)
  • Code explanation — explain what a block of code does
  • Code transformation — modernise Java 8/11 → Java 17/21 (Java upgrade feature)
  • Security scanning — detect vulnerabilities in code (SAST-like)
  • Unit test generation — generate test cases for selected code
  • CLI companion/dev command in AWS CloudShell; Q explains AWS CLI commands
  • Console integration — asks Q about services in the AWS Management Console

Integrations

  • IDE plugins: VS Code, JetBrains IDEs (IntelliJ, PyCharm, etc.), Visual Studio
  • AWS Console: Q button in every service page
  • AWS CloudShell: /dev natural language to AWS CLI translation
  • Amazon CodeCatalyst: integrated throughout the dev workflow

Amazon Q Developer vs GitHub Copilot

DimensionQ DeveloperGitHub Copilot
AWS-awarenessDeep — understands AWS services, IAM, CDKGeneric
Security scanningBuilt-in SASTVia third-party extensions
Java modernisationAutomated Java 8→17 upgradeManual
PricingFree tier (50 inline/month); Pro $19/user/month$10/user/month
IDE supportVS Code, JetBrains, Visual StudioVS Code, JetBrains, Neovim, etc.

Exam trigger for Q Developer: "AI coding assistant in the IDE", "generate code for AWS services", "explain code in VS Code", "detect security vulnerabilities in code", "Java modernisation"


Exam Scenario Drill

ScenarioService
Employees ask HR questions and get paragraph answers from ConfluenceQ Business
Employees search for documents about a policy and see a list of matching passagesKendra
Developer asks for code completion while writing a Lambda functionQ Developer
Company wants a GenAI assistant with SharePoint ACL enforcementQ Business
Developer wants security vulnerability scanning inside VS CodeQ Developer
Company wants ML-ranked search results from 10 internal data sourcesKendra

Key Facts

  • Amazon Q Business = GenAI assistant over company data; synthesises answers with citations; permission-aware via ACL enforcement
  • Amazon Kendra = enterprise ML search; returns ranked document passages; does not generate answers
  • Amazon Q Developer = AI coding assistant; IDE plugins (VS Code, JetBrains); inline completion + security scanning + Java modernisation
  • Q Business uses Bedrock FMs under the hood; Kendra uses ML ranking models (not generative)
  • Q Business plugins allow actions in third-party systems (Jira, Salesforce) from within the assistant
  • Q Developer free tier: 50 inline suggestions/month; Pro tier: $19/user/month

Connections

Open Questions

  • Is Amazon Q Business converging with Bedrock Agents, or are they on separate roadmaps?
  • Does Q Business's ACL enforcement work at query time (re-checks permissions on every query) or at index time only?