How shreyaasdeworld/ai_hiring_intelligence_platform works
This is a Streamlit resume-screening application. A recruiter enters job requirements and uploads a resume; the app validates inputs, extracts text, estimates token usage, chunks the resume, and stores its embeddings in ChromaDB. It then sends the resume and requirements through a LangChain prompt/model/parser chain, displays the screening report, and offers a download. The README also describes semantic retrieval as a capability, but the shown app stores embeddings without performing retrieval. All internal components are implemented in app.py; external API links are not directly shown.
Main entry points: app.py Input validation
Components 18
Recruiter workflow
- Streamlit UI
app.py - Input validation
app.py - Report display
app.py - Report download
app.py
Resume processing
- Resume extraction
app.py - PDF reader
app.py - DOCX reader
app.py - TXT reader
app.py - Token counting
app.py - Text chunking
app.py - Embedding client
app.py - ChromaDB store
app.py
AI screening
- Screening prompt
app.py - LangChain chain
app.py - Chat model
app.py - Output parser
app.py
Other components
- Recruiter
- OpenAI API
Connections 18
- Recruiter → Streamlit UI: enters inputs
app.py - Streamlit UI → Input validation: validates inputs
app.py - Input validation → Resume extraction: extracts resume
app.py - Resume extraction → PDF reader: reads PDF
app.py - Resume extraction → DOCX reader: reads DOCX
app.py - Resume extraction → TXT reader: reads TXT
app.py - Input validation → Token counting: counts tokens
app.py - Input validation → Text chunking: splits text
app.py - Input validation → ChromaDB store: stores chunks
app.py - ChromaDB store → Embedding client: embeds chunks
app.py - Input validation → LangChain chain: invokes screening
app.py - LangChain chain → Screening prompt: formats prompt
app.py - LangChain chain → Chat model: calls model
app.py - LangChain chain → Output parser: parses output
app.py - Input validation → Report display: shows report
app.py - Report display → Report download: offers download
app.py - Embedding client → OpenAI API: requests embeddings
- Chat model → OpenAI API: requests completion