How haq-nawaz-malik/agentic_resume works
Project Type and Purpose
• AgenticResume is a full-stack web application that automates resume tailoring and interview prep by orchestrating multiple AI agents.
• Frontend: React + Tailwind CSS UI for user auth, resume/job management, and dashboard.
• Backend: Flask REST API with SQLAlchemy for data modeling and JWT for auth.
• AI Component: CrewAI-based multi-agent system (researcher, profiler, strategist, preparer).
• Database: PostgreSQL.
• External Services: OpenAI API for language tasks, Serper for web search.
File Structure Analysis
• Top Level:
– ai/ → AI agents & tasks & tooling (file scraping, web search).
– backend/ → Flask app, config, models, routes, services, utils.
– frontend/ → React SPA structure (components, pages, services, context).
– tests/ → pytest test suites for auth and resume flows.
– Dockerfile(s) & docker-compose.yml → container definitions for local/dev setup.
• Architectural Patterns:
– Layered backend (models → services → routes).
– Componentized frontend (layout, UI, forms, pages, context).
– Micro-service-style isolation between AI logic (ai/) and core API (backend/).
Key README Insights
• Four specialized AI agents collaborating via CrewAI.
• Explicit callouts of tech stack and deployment with Docker Compose.
a. Main Containers:
1. Frontend (React SPA + Nginx)
2. Backend API (Flask + SQLAlchemy + JWT)
3. AI Agents Module (CrewAI orchestration)
4. Database (PostgreSQL)
5. External Services (OpenAI, Serper)
b. Relationships and Data Flows:
– User ↔ Frontend: HTTP/HTTPS (login, dashboard, resume builder)
– Frontend → Backend API: REST calls via api.js service layer
– Backend → Database: SQL over TCP (via SQLAlchemy)
– Backend → AI Module: gRPC/HTTP or Python import calls to ai_service.py
– AI Module → OpenAI API & Serper API: HTTPS requests for LLM and web search
– Backend → Frontend: REST JSON responses
– User → Export: PDF/Markdown generation handled by backend or AI module
c. Architectural Patterns & Principles:
– Separation of concerns: UI, business logic, data, and AI logic are decoupled.
– Layered modularity in backend (routes → services → models).
– Component and context pattern in frontend for state management.
– Dependency inversion: ai_service abstracts AI interactions behind a service interface.
d. Key Technologies & Frameworks:
– Frontend: React, Tailwind CSS, Context API, Jest for tests
– Backend: Flask, Flask-JWT-Extended, SQLAlchemy, pytest
– AI Module: CrewAI framework, OpenAI Python client
– DevOps: Docker, Docker Compose, Nginx