Spring RAG API

Spring Boot 3 RAG service using LangChain4j, Ollama (chat + embeddings), and H2 for vector storage.

Prereqs

• Java 17+
• Maven
• Ollama running locally (default http://localhost:11434)
  • Pull models: ollama pull llama2 (or mistral), ollama pull nomic-embed-text

Run

mvn spring-boot:run
# or
mvn clean package && java -jar target/spring-rag-app-0.0.1-SNAPSHOT.jar

H2 file DB is at ./data/ragdb (console at /h2-console, user sa, password password).

Config

Edit src/main/resources/application.yml:

• rag.ollama.base-url, rag.ollama.model (chat), rag.ollama.embedding-model
• Chunking: rag.chunking.max-tokens, rag.chunking.overlap
• Retrieval: rag.top-k

REST Endpoints

• POST /api/documents (multipart file): ingest PDF/TXT, auto-embeds chunks
• GET /api/documents : list docs
• GET /api/documents/{id} : doc metadata
• GET /api/documents/{id}/chunks : chunks
• POST /api/rag/query (JSON {question, topK}) : RAG answer with context

Sample Requests/Responses

Ingest (multipart)

curl -X POST http://localhost:8080/api/documents \
  -F "file=@/path/to/sample.txt"

Response:

{
  "documentId": "4b7e93c0-1c9f-4e9f-9b02-2e06d3e6f6c2",
  "chunksIndexed": 3
}

List documents

curl http://localhost:8080/api/documents

Response:

[
  {
    "id": "4b7e93c0-1c9f-4e9f-9b02-2e06d3e6f6c2",
    "title": "sample",
    "originalFileName": "sample.txt",
    "contentType": "text/plain",
    "sizeBytes": 1200,
    "createdAt": "2025-12-11T21:00:00Z",
    "updatedAt": "2025-12-11T21:00:00Z",
    "chunkCount": 3
  }
]

Get chunks for a document

curl http://localhost:8080/api/documents/4b7e93c0-1c9f-4e9f-9b02-2e06d3e6f6c2/chunks

Response:

[
  {
    "id": "7d1f7a3e-5bf7-4b71-8a7d-65b42d5f8e93",
    "chunkIndex": 0,
    "text": "Chunk text ..."
  },
  {
    "id": "b3e8949f-1e1d-4e62-8bb1-3b5c9a5f2b1e",
    "chunkIndex": 1,
    "text": "Chunk text ..."
  }
]

RAG query

curl -X POST http://localhost:8080/api/rag/query \
  -H "Content-Type: application/json" \
  -d '{"question":"What does the sample document say?","topK":3}'

Response:

{
  "question": "What does the sample document say?",
  "answer": "It explains the content of the sample document ...",
  "context": [
    "Chunk text ...",
    "Another chunk ..."
  ]
}

Quick HTTP Tests (IntelliJ HTTP Client)

Use src/test/resources/http/rag-api.http and run requests against http://localhost:8080.

1. Ingest sample text
2. List docs, capture id
3. Get chunks
4. POST RAG query

Notes / Troubleshooting

• Ensure Ollama server is up and the embedding model exists (ollama list).
• If H2 lock issues, stop the app and delete data/ locally (DB will re-init). The folder is gitignored.
• Logs set to warn for SQL; enable debug with --debug if needed.