How ekas-7/zing-guru works
Project type and purpose
• Zing Guru is a full-stack, AI-driven, gamified learning platform that integrates:
• A React-based web frontend (MERN stack)
• Node.js microservices (Notion-Planner service + core API server)
• Python-based ML/GenAI microservices (RAG service, Video Summarizer)
• A Chrome extension for quick access
• External systems: MongoDB Atlas, Ethereum/Polygon blockchain, TensorFlow models
File structure
• Top-level “Frontend” directory: React/Vite application with Tailwind CSS, components, pages, config, store.
• “Backend” directory:
• Notion-Planner (Node.js service for saving user notes to Notion)
• server (Node.js/Express core API handling YouTube transcripts)
• “ML-APi” directory: two Python microservices: RAG (FastAPI) and Video-summarizer (Flask or FastAPI)
• “extension”: Chrome extension assets and scripts
• Root config files: package.json, monorepo–style dependencies
• README outlines blockchain integration, AI/ML, and overall architecture
README for additional insights
• Tech stack confirms: MERN frontend, FastAPI backend for core API and ML, TensorFlow, Ethereum rewards, MongoDB Atlas
• Key features: AI personalization (calls to ML microservices), gamification (NFT minting via Web3), accessibility (voice assistant)
Main components and their relationships
a. Frontend (React app)
• Hosts UI: Landing, Login, Dashboard, Notion editor, VideoSummarizer, VoiceAssistant
• Calls Core API (Node/server) for authentication, user data, transcript processing
• Calls Notion-Planner service for note saving
• Dispatches actions to Redux store, communicates with Firebase (if used for auth) and Navigation config
b. Core API Server (Backend/server)
• Exposes REST endpoints for YouTube transcript fetch & summarization
• Interacts with MongoDB Atlas for user profiles, progress, credentials
• Calls ML microservices (RAG & Video Summarizer) via HTTP
• Bridges frontend requests to blockchain modules (smart contract interactions)
c. Notion-Planner Microservice
• Node.js service that authenticates with Notion API and saves user notes
• Invoked from frontend Dashboard “Notion” component
d. ML Microservices (ML-APi)
• RAG service: receives queries, loads retrieval models, returns AI-driven content
• Video Summarizer: processes video URLs, returns transcript or summary
e. Chrome Extension
• Lightweight UI in browser popup
• Can send URLs or text to Core API or ML services for on-the-fly summarization
f. External systems
• MongoDB Atlas: stores user profiles, paths, progress, credentials
• Ethereum/Polygon blockchain: mints NFTs via Web3 calls (could be embedded in Core API or frontend)
• TensorFlow models loaded in ML services or via external AI endpoints
Key architectural patterns and design principles
• Microservices: separate Node.js services (Core API, Notion-Planner) and Python services (RAG, Summarizer)
• Event-driven or request/response between services via HTTP REST
• Single Page Application (SPA) with API gateway (Core API)
• External integrations (Notion, Blockchain, Firebase, MongoDB Atlas)
• Redux for state management, Tailwind for styling, Vite bundler
Essential details to capture
• Authentication & authorization flow (Firebase or JWT)
• AI-powered endpoints (highlight RAG & summarizer calls)
• Blockchain reward minting path (frontend ↔ Core API ↔ Web3 ↔ Ethereum)
• Notion integration path
• State management boundary (Redux store)