How boubacar593/rag_agentrecruteur works
Recruteur Augmente AI is a local RAG assistant for recruiters: upload or edit a job description, index PDF CVs, then ask questions to compare retrieved candidate profiles against the role. The graph separates UI orchestration, document preparation, retrieval, and local model inference. README capabilities are represented, but the index and sampled code do not show a separate ranking engine or privacy enforcement beyond the stated local Ollama design.
Main entry points: app.py Recruitment UI · chroma_db Chroma vector store · config.py Runtime configuration · rag.py RAG chain
Components 12
Recruiter interface
- Recruitment UI
src/app.py
Document preparation
- Job upload extraction
src/utils.py - CV PDF loading
src/ingestion.py - Text chunking
src/ingestion.py
RAG analysis
- RAG chain
src/rag.py - Candidate retrieval
src/rag.py - Role-profile comparison
src/rag.py
Local AI and storage
- Runtime configuration
src/config.py - Embedding model
src/database.py - Chroma vector store
database/chroma_db - Local chat model
src/rag.py
Other components
- Recruiter
Connections 17
- Recruiter → Recruitment UI: uses
README.md - Recruitment UI → Job upload extraction: extracts job text
src/app.py - Recruiter → Recruitment UI: requests CV indexing
src/app.py - Recruitment UI → CV PDF loading: loads CVs
src/app.py - CV PDF loading → Text chunking: splits documents
src/ingestion.py - Recruitment UI → Chroma vector store: stores chunks
src/app.py - Recruitment UI → RAG chain: creates and invokes
src/app.py - RAG chain → Candidate retrieval: searches profiles
src/rag.py - Candidate retrieval → Chroma vector store: similarity search
src/rag.py - RAG chain → Role-profile comparison: builds prompt
src/rag.py - Role-profile comparison → Local chat model: invokes
src/rag.py - Recruitment UI → Embedding model: configures store
src/database.py - Embedding model → Chroma vector store: embeds documents
src/database.py - Runtime configuration → Chroma vector store: sets persistence path
- Runtime configuration → Embedding model: sets model
src/database.py - Runtime configuration → Local chat model: sets model
src/rag.py - Recruitment UI → Recruiter: shows answer
src/app.py