ujk768/rag-pipeline-backend

How ujk768/rag-pipeline-backend works

This is a local PDF RAG API. Users upload a PDF, monitor background ingestion, then query it for grounded answers with page citations; they can also inspect pruning results or reset stored data. The graph preserves ingestion, retrieval, generation, pruning, and operational endpoints. The README documents the query pipeline, while the sampled main.py excerpts show database setup, ingestion details, pruning, and API behavior; unsampled query internals are represented only where the README supports them.

Main entry points: main.py FastAPI endpoints

Components 15

API and Operations

FastAPI endpoints main.py
Processing status main.py
Reset data main.py

PDF Ingestion

Retrieval and Answers

Query embedding main.py
Similarity search main.py
Local LLM generation main.py
Cited answer main.py

Chunk Pruning

Pruning strategies main.py
Pruning report main.py

Other components

User
PostgreSQL pgvector

Connections 18

User → FastAPI endpoints: uploads or queries README.md
FastAPI endpoints → PDF extraction: starts background ingestion main.py
PDF extraction → Sentence chunking: extracts and chunks README.md
Sentence chunking → Document embeddings: embeds chunks README.md
Document embeddings → Pruning strategies: optionally prunes main.py
Pruning strategies → Chunk vector store: stores kept chunks README.md
Document embeddings → Chunk vector store: stores all chunks README.md
Chunk vector store → PostgreSQL pgvector: persists vectors main.py
FastAPI endpoints → Processing status: serves progress main.py
FastAPI endpoints → Pruning report: serves report main.py
Pruning strategies → Pruning report: builds statistics main.py
FastAPI endpoints → Query embedding: handles query README.md
Query embedding → Similarity search: provides query vector README.md
Similarity search → Chunk vector store: retrieves top chunks README.md
Similarity search → Local LLM generation: supplies context README.md
Local LLM generation → Cited answer: generates grounded answer README.md
FastAPI endpoints → Reset data: handles reset main.py
Reset data → Chunk vector store: deletes chunks main.py

More Python diagrams: project-lighter/lighter · renatoalmeidaoliveira/nbrisk · rustedpy/maybe · technikum-teb-edukacja-we-wroclawiu/inf.04-rozwiazania · wchao0601/m4-sar

Drawn by GitDiagram on September 30, 2026 from the repository's file tree, README and source files. Read as Markdown ·

How ujk768/rag-pipeline-backend works

Main entry points: main.py FastAPI endpoints

Connections 18

User → FastAPI endpoints: uploads or queries README.md
FastAPI endpoints → PDF extraction: starts background ingestion main.py
PDF extraction → Sentence chunking: extracts and chunks README.md
Sentence chunking → Document embeddings: embeds chunks README.md
Document embeddings → Pruning strategies: optionally prunes main.py
Pruning strategies → Chunk vector store: stores kept chunks README.md
Document embeddings → Chunk vector store: stores all chunks README.md
Chunk vector store → PostgreSQL pgvector: persists vectors main.py
FastAPI endpoints → Processing status: serves progress main.py
FastAPI endpoints → Pruning report: serves report main.py
Pruning strategies → Pruning report: builds statistics main.py
FastAPI endpoints → Query embedding: handles query README.md
Query embedding → Similarity search: provides query vector README.md
Similarity search → Chunk vector store: retrieves top chunks README.md
Similarity search → Local LLM generation: supplies context README.md
Local LLM generation → Cited answer: generates grounded answer README.md
FastAPI endpoints → Reset data: handles reset main.py
Reset data → Chunk vector store: deletes chunks main.py

Drawn by GitDiagram on September 30, 2026 from the repository's file tree, README and source files. Read as Markdown ·