sinapars9/doclens

How sinapars9/doclens works

DocLens is a Streamlit PDF question-answering assistant. Users upload and select documents, the system parses and indexes them, then resolves the question, retrieves and reranks evidence, streams an answer, and verifies its claims. The graph preserves documented multi-document scope, parent-child retrieval, caching, and human feedback. The pipeline excerpt omits some orchestration calls, so documented links use the README; unsampled implementation details are not inferred.

Main entry points: pipeline.py RAG orchestrator · chunk_pdf.py Semantic chunks · vector_store.py Vector store · query_rewriter.py Query resolution

Components 18

PDF ingestion

PDF extraction inspect_pdf.py
Semantic chunks chunk_pdf.py
ONNX embeddings embedding.py
Chunk context contextualizer.py
Model settings config.py

Search and context

Vector store vector_store.py
Hybrid retrieval retriever.py
Query resolution query_rewriter.py
Cross-encoder reranking reranker.py
Semantic cache semantic_cache.py

Answer and verification

RAG orchestrator pipeline.py
Streaming generation generate.py
Answer verifier verifier.py

User interface

Streamlit interface app.py
Feedback log app.py

Other components

Document user
LLM service
Disk cache

Connections 22

Document user → Streamlit interface: uploads and asks README.md
Streamlit interface → RAG orchestrator: submits questions README.md
Streamlit interface → Feedback log: logs feedback app.py
RAG orchestrator → Semantic chunks: builds document index README.md
Semantic chunks → PDF extraction: loads PDF content chunk_pdf.py
Semantic chunks → ONNX embeddings: embeds semantic windows chunk_pdf.py
Semantic chunks → Chunk context: contextualizes chunks chunk_pdf.py
Semantic chunks → Vector store: stores indexed chunks README.md
Vector store → Semantic chunks: builds from chunks vector_store.py
Vector store → ONNX embeddings: embeds stored text vector_store.py
RAG orchestrator → Query resolution: resolves query pipeline.py
Query resolution → LLM service: requests query expansion query_rewriter.py
Query resolution → Disk cache: reads and writes query_rewriter.py
RAG orchestrator → Semantic cache: checks semantic cache pipeline.py
RAG orchestrator → Hybrid retrieval: retrieves candidates README.md
Hybrid retrieval → Vector store: searches vectors retriever.py
RAG orchestrator → Cross-encoder reranking: reranks candidates pipeline.py
RAG orchestrator → Streaming generation: streams answer README.md
Streaming generation → LLM service: requests completion generate.py
Streaming generation → Disk cache: reads and writes generate.py
RAG orchestrator → Answer verifier: verifies answer README.md
ONNX embeddings → Model settings: loads model path embedding.py

Drawn by GitDiagram on November 13, 2026 from the repository's file tree, README and source files. Read as Markdown ·