How rxz33/ai-interview-generator works
Project type + purpose (what you’re diagramming)
This is a full-stack web application: InterQ (AI Interview Question Generator). Users interact with a React + Tailwind frontend, which calls a Node.js/Express backend. The backend generates interview questions and sample answers by calling Groq AI, and persists data in MongoDB.
Given the file tree:
• client/ → frontend (React app)
• server.js → backend entrypoint (Express server)
• package.json / package-lock.json → likely root Node project (possibly serving API + building client)
Frontend (client/)
• UI Pages / Views
• Role selection, difficulty selection, skill level, company type, etc.
• API Client Layer
• A thin module responsible for calling backend endpoints (fetch/axios)
• State Management
• Local state / context / query cache (don’t assume Redux; label generically)
• Styling
• Tailwind CSS (call out as a frontend dependency)
Backend (server.js)
Treat server.js as the entrypoint that wires:
• Express App / Router layer
• Interview Generation Controller / Service
• Builds prompts, selects difficulty, formats output
• Groq AI Client
• Outbound API calls (LLM inference)
• MongoDB Data Access Layer
• Config / Secrets
Deployment (render.yaml)
An infra boundary that shows:
• Render Web Service hosting Node/Express
• MongoDB (likely MongoDB Atlas) as an external managed service
• Groq API as an external SaaS dependency
User Browser
→ loads React SPA (from CDN/static hosting or from Express static serving)
React SPA
→ calls Express REST API (HTTPS)
Express API
→ reads/writes MongoDB (TCP connection via driver/ODM)
Express API
→ calls Groq AI API (HTTPS)
Express API
→ returns structured JSON to React SPA
React SPA
→ renders question set + sample answers + follow-ups
• Does Express serve the React build or is React deployed separately?
• Actual API routes (e.g., /api/generate, /api/questions, /api/history)
• Mongo usage: which entities are stored (history, users, templates)
• render.yaml: single service vs multiple services (web + static)
This will be the most accurate representation of this repo’s purpose and structure while still being detailed and componentized.