How himanshu-skid19/pathfinder-ai-enhanced-resumes works
Project type and purpose
• From the README and file tree, this is a single-codebase AI-powered resume parsing and cover-letter generation tool with a Streamlit UI (“app.py”), backed by a FAISS vector store for semantic search, and templating modules that render LaTeX outputs. It is not a multi-service microservice topology but a monolithic “full-stack” data-science application.
File structure
• Top-level files:
• app.py – Streamlit front-end and orchestration
• create_vector_store.py – indexing resumes into FAISS
• read_data.py – parsing and preprocessing resumes
• prompts.py, llm.py – prompting logic and OpenAI API wrapper
• iitg_template.py, latex_template.py – rendering templates
• imports.py – shared imports/config
• test_data.py – sample or unit tests
• requirements.txt – dependencies (streamlit, faiss, openai, etc.)
• resumes/ – raw PDF resumes served as data
README for additional insights
• It clarifies prerequisites (WSL for FAISS), the launch steps, and a brief “Project Architecture” listing the same files.
Key details to capture
• The conditional step: vector store build vs. reuse
• The synchronous LLM call and templating sequence
• The final output artifact (LaTeX/PDF)