AI service
Amazon Kendra
Enterprise semantic search across connected content repositories.
Key points
- Provides intelligent enterprise search over connected content repositories.
- Uses natural language understanding so users can ask questions instead of only keyword queries.
- Supports connectors to common enterprise data sources and respects access controls where configured.
- Returns ranked search results and answers from indexed content.
- Targets enterprise search rather than text classification or foundation model inference.
When to use it
- Choose Kendra when employees need semantic search across enterprise documents and repositories.
- Use it for intranet search, support knowledge bases, policy search, or technical documentation search.
- Use it when the question emphasizes connectors, indexing, and search relevance.
Exam tips
- Kendra is enterprise search; Bedrock Knowledge Bases are Bedrock-native RAG for foundation model grounding.
- Comprehend extracts insights from text, while Kendra helps users find answers across repositories.
- OpenSearch is a search engine service; Kendra is the managed intelligent search AI service.
- Access control is important so search results do not expose restricted documents.