Case study: developer education
A technical-books catalogue with context built in.
Browse authors and topics, search chapter excerpts by meaning, and keep the source book attached to every recommendation.
Flat chunk search loses the catalogue around the answer.
A useful book explorer should answer a question and show where to read next. Chunk-only retrieval can find an excerpt while dropping the author, title, topic, and chapter provenance needed for a trustworthy recommendation.
Before
- Book metadata and chunks queried separately
- Recommendations return isolated excerpts
- Author and topic facets need manual joins
- Search UI hard-codes every available filter
With RushDB
- Books, authors, topics, chapters, and chunks stay connected
- Chapter text is searchable by meaning
- Recommendations retain source provenance
- Ontology can inform available browse filters
Graph intelligence on ingest
Incoming data becomes queryable graph context.
RushDB turns structured data into graph-ready context without a separate modeling pipeline. Structure already encoded in a nested payload is linked immediately. For flat records imported from scattered sources, relationship analysis can propose stable cross-source patterns.
01
Normalize as data arrives
Import JSON or CSV. RushDB infers property types and adds new fields to the live, queryable ontology without a schema migration.
02
Auto-link nested structure
Nested objects become connected records automatically, preserving the parent-child graph structure already encoded in your payload.
03
Enrich scattered sources
After flat imports or schema changes, analyze the project ontology. RushDB can suggest join patterns and semantic relationship types for your review.
Suggested relationship analysis requires an LLM configured for the project. Suggestions stay in draft form until you approve them, so inferred domain meaning never mutates the graph silently. You can also add explicit relationships through the SDK or API.
Review suggested relationship patternsData model
One flexible graph for the workflow.
Start with the payload shape your product already produces. RushDB stores it as Records, infers typed properties, and keeps nested or approved domain relationships queryable.
Catalogue metadata, chunks, authors, and topics stay connected so retrieval can return provenance.
{
"isbn": "978-1-0000-0042-1",
"title": "Reliable Services in Practice",
"AUTHOR": [{ "name": "Mina Patel" }],
"TOPIC": [{ "name": "distributed systems" }, { "name": "backpressure" }],
"CHAPTER": [{
"title": "Backpressure and recovery",
"chapterNumber": 7,
"BOOK_CHUNK": [
{ "chunkId": "chunk-7-03", "text": "Bound queues keep downstream failure visible..." }
]
}]
}Working example
Recommend a book and cite the chapter.
Search indexed chapter excerpts for a practical distributed-systems question, then return the matching book, chapter, author, and topic context.
BOOK "Reliable Services in Practice"
AUTHOR "Mina Patel"
TOPIC "distributed systems"
CHAPTER "Backpressure and recovery"
BOOK_CHUNK.text "Bound queues keep downstream failure visible..."{
"labels": ["BOOK_CHUNK"],
"propertyName": "text",
"query": "practical tracing, backpressure, and recovery patterns"
}{
"book": "Reliable Services in Practice",
"author": "Mina Patel",
"chapter": "Backpressure and recovery",
"topic": "distributed systems"
}TypeScript SDK
A complete starting point.
The example includes the one-time index setup so the retrieval path is explicit.
Implementation blueprint
Build the books GraphRAG path.
Use this sequence to keep catalogue browsing, semantic retrieval, and citation provenance on the same graph.
- 01Import BOOK records with nested AUTHOR, TOPIC, CHAPTER, and BOOK_CHUNK records
- 02Create a managed index for BOOK_CHUNK.text
- 03Retrieve chunks by semantic fit and optional topic filters
- 04Enrich matching chunks with connected book, chapter, author, and topic records
- 05Evaluate retrieval against a small question set before expanding the catalogue
Build path
- Keep BOOK_CHUNK records connected to CHAPTER, BOOK, AUTHOR, and TOPIC records.
- Use ontology to expose available browse filters.
- Return citations and provenance with each recommendation.
- Treat chunking and public-data backfill as application concerns.
Relevant docs
Read the exact primitives behind this pattern.
These links point to the RushDB docs pages that map directly to this case study: ingestion, labels, properties, values, SearchQuery, relationships, semantic search, MCP, or deployment.
Semantic search
Search book chunks by meaning while keeping chapter, author, topic, and citation context connected.
Open docsEmbedding indexes
Create and monitor embedding indexes for searchable text properties such as BOOK_CHUNK.text.
Open docsRelationships
Connect chunks to books, chapters, authors, topics, and recommendations for provenance-aware GraphRAG.
Open docsHow it works
Build the smallest useful workflow first.
01
Import the catalogue graph
Store books with nested authors, topics, chapters, and chunks so the browse model and retrieval model share one substrate.
02
Index chapter excerpts
Create one managed index for chunk content and search the catalogue with natural-language questions.
03
Return provenance
Enrich matching chunks with their connected chapter, book, author, and topic records before assembling an answer.
Know where it fits.
Start from a narrow golden path
Ship one browse flow and one cited recommendation flow before expanding the catalogue UI.
Evaluate retrieval explicitly
Maintain a small ground-truth set of questions and measure retrieval quality as chunking or ranking changes.
Questions developers ask.
Next step