How manishjadhav9/vikasini works
Project Type & Purpose
• This is a full-stack web application (“AI Upskilling Platform for Underprivileged Women”) that combines a Next.js frontend (User + Admin), a Node-based backend with SQLite/Prisma, and multiple AI/third-party integrations.
• Its goal is to provide multilingual, voice-enabled learning, AI-driven personalized paths, gamification, job matching, and an admin dashboard for content/stats.
File‐Structure Analysis
Top Level
• README.md, package.json
• vikasini-backend/ – REST API service
– src/controllers, src/routes, src/models, src/db
– SQLite database file (data/vikasini.db)
– Scripts (create-controller.js, fix-models.js, show-users.js)
• vikasini-frontend/ – Next.js 13 “app” directory
– app/ (public pages + API routes under app/api)
– components/ (UI elements, layout, theme)
– hooks/, lib/ (contexts, utilities, API client)
– data/ (simulated JSON learning-paths & progress)
– config (next.config.js, tailwind, postcss)
Patterns
• Frontend: Next.js App Router, ISR/SSR, built-in API routes for mock/proxy
• Backend: MVC structure (controllers ↔ models via Prisma ORM), Express-style routing
• Shared: SQLite for persistence, Prisma for schema & queries
README Insights
• Frontend: Next.js + Tailwind + Framer Motion
• Backend: SQLite + Prisma ORM
• Auth: NextAuth.js or Clerk
• AI/ML: OpenAI GPT, TinyLLaMA/Ollama, FAISS, VADER/TextBlob
• Speech/Translation: Sarvam API (STT/TTS)
• File storage: Cloudinary/Firebase/Supabase
• Visualization: Chart.js/ECharts