Chunking, search indexes and retrievers in Data 360
How Data 360 makes unstructured content retrievable: chunking and vectorization, vector, hybrid and enriched search indexes, data-space scoping, and individual and ensemble retrievers with filters, versions, citations and the Retriever Playground.
Explain foundational concepts of Data 360 such as chunking, indexing, and retrievers.
~7 min read · 32 practice questions ·
What the exam expects
- Explain how unstructured content becomes searchable: ingestion, mapping, chunking, vectorization and the search index.
- Choose a vector or hybrid search index, and decide when enrichment or image processing is worth it.
- Configure an individual retriever: filters and filter logic, returned fields and result count, versions, citations, and the Retriever Playground.
- Decide when an ensemble retriever is needed and know its limits.
- Describe the run-time sequence: retrieve, augment the prompt, generate.
Key terms: chunk is a semantically meaningful passage of a document; vector embedding is a numeric representation of a chunk's meaning; search index stores chunked, vectorized content for one data model object; individual retriever searches one index for one use case; ensemble retriever runs several individual retrievers and reranks the combined results.
Key facts
- Detail
- Defined in a data space, associated with one DMO
- Detail
- Vector, hybrid and enriched; enriched chunks are turned on when creating a vector or hybrid index
- Detail
- Consumes Flex Credits
- Detail
- Ten filter conditions per individual retriever
- Detail
- Each save creates a version; one active at a time
- Detail
- Enabled per individual retriever in AI Models
- Detail
- Enable image processing on the unstructured DMO search index
- Detail
- An activated retriever is available to Prompt Builder users
Sources for this section (4)
- Why does a hybrid index return a document containing an exact part number that a vector index missed?
- When is an ensemble retriever the right choice?
- Which three chunks does an enriched index generate per passage?
- Why is content chunked before indexing?
The questions are free to read. Answers with their sources and self-grading are included with the pack.
Keep reading with the Agentforce Specialist pack
- Core concepts
- Exam traps
- Worked scenario
The pack includes:
- All 26 lessons of the Agentforce Specialist guide, with reading progress
- Interactive decision walks and step-by-step sequence steppers
- Inline practice questions and one-click topic practice (32 questions for this lesson)
- An official source cited for every claim
- An AI tutor that answers from the guide
- Unlimited timed practice exams