AI-Powered Personalized Learning Path Recommender, Milestone Roadmap Generator, and Skill Mastery Platform.
- Application URL: https://neuronpath-ai.vercel.app/
NeuronPath AI is an intelligent, full-stack career acceleration and learning roadmap platform. It takes a learner's natural language career ambition, analyzes baseline proficiencies against required target skills, synthesizes custom prerequisite-ordered milestone roadmaps, curates multi-format learning resources, and adapts dynamically through technical assessments and conversational AI coaching.
- 🗺️ Personalized Learning Roadmaps: Dynamically generates fine-grained milestones decomposed with strict topological prerequisite ordering based on the user's chosen track.
- 🗣️ AI-Powered Onboarding: Conversational goal extraction powered by Google Gemini LLM to automatically parse target roles, timelines, and known skills.
- 🎯 Career & Learning Goals: Target role alignment with estimated timelines, proficiency benchmarks, and personalized milestones.
- 📊 Skill Gap Analysis: Real-time evaluation comparing current learner proficiencies with target industry requirements.
- 📚 Curated Learning Resources: Verified courses, official documentation, video masterclasses, interactive coding sandboxes, and canonical industry textbooks mapped to specific milestones.
-
📝 Assessments & Progress Tracking: Milestone-aligned quizzes with automated scoring, adaptive feedback (
$\ge 70%$ mastery threshold), and comprehensive visual analytics. - 🤖 Context-Aware AI Learning Coach: Real-time conversational guidance with live learner context injection (active goal, current milestone, and skill gaps).
- 🔀 Multi-Goal Management & Switching: Supports up to 3 distinct learning goals with instant 1-click switching, isolated roadmap progression, and safe goal deletion.
- Frontend: React 18, TypeScript, Tailwind CSS, Vite, Zustand (State Management), Lucide React (Icons), Recharts (Visualizations), Axios (API Client)
- Backend: FastAPI, Python 3.10+, SQLAlchemy ORM, Pydantic v2, Uvicorn, PyJWT (Authentication), bcrypt (Password Hashing)
- Database: PostgreSQL (Supabase in production) / SQLite (Local development)
- AI / LLM: Google Gemini API (
gemini-2.0-flashviagoogle-generativeai), with provider fallback architecture - Deployment: Vercel (Frontend SPA), Render (Backend API Service)
┌────────────────────────────────────────────────────────┐
│ Frontend (Client) │
│ React 18 • TypeScript • Tailwind CSS • Vite │
│ Zustand (State) • Lucide Icons • Axios │
└───────────────────────────┬────────────────────────────┘
│ REST APIs (/api/*)
▼
┌────────────────────────────────────────────────────────┐
│ Backend (Server) │
│ FastAPI • Uvicorn • Pydantic v2 │
│ SQLAlchemy ORM • Directed Acyclic Graph │
└─────────────┬────────────────────────────┬─────────────┘
│ │
▼ ▼
┌───────────────────────┐ ┌─────────────────────────┐
│ Database Storage │ │ LLM Intelligence │
│ PostgreSQL / SQLite │ │ Google Gemini API │
└───────────────────────┘ └─────────────────────────┘
- Sign Up & Onboarding: The user signs up and enters their learning ambition in natural language.
- Goal Extraction: The AI extracts the target role, experience level, and timeline, creating their initial goal.
- Roadmap Generation: A Directed Acyclic Graph (DAG) organizes sequential milestones with curated learning resources.
- Learning & Milestone Assessments: The learner works through milestone items and completes technical quizzes to validate mastery and unlock next steps.
- Multi-Goal Switching & AI Coach: Learners can add up to 3 distinct goals, switch active paths seamlessly, and chat with the AI coach for contextual guidance.
- Node.js 18+ & npm
- Python 3.10+
cd backend
# Create virtual environment
python -m venv venv
# Activate virtual environment
# Windows:
venv\Scripts\activate
# macOS/Linux:
source venv/bin/activate
# Install dependencies
pip install -r requirements.txt
# Run FastAPI backend server
uvicorn main:app --host 127.0.0.1 --port 8000 --reloadThe backend API will be available at http://127.0.0.1:8000 (Interactive Swagger Docs at /docs).
cd frontend
# Install dependencies
npm install
# Run Vite dev server
npm run devThe frontend application will be accessible at http://localhost:5173.
- Frontend: Hosted on Vercel (https://neuronpath-ai.vercel.app/)
- Backend: Hosted on Render (https://neuronpath-api.onrender.com/)
The deployed backend runs on a free-tier service and may enter a sleep state. The first request after inactivity can take approximately 60 seconds to wake up. Subsequent requests should respond normally.
NeuronPath AI/
├── backend/
│ ├── ai/ # LLM provider integration & onboarding extraction
│ ├── api/ # FastAPI routers (auth, goals, roadmap, skills, etc.)
│ ├── core/ # Security, JWT, and authentication dependencies
│ ├── models/ # SQLAlchemy ORM models
│ ├── recommendation/ # Prerequisite graph and recommendation scoring
│ ├── schemas/ # Pydantic validation schemas
│ ├── services/ # Business logic (goals, roadmaps, assessments)
│ ├── tests/ # Pytest test suites
│ ├── main.py # Application entrypoint & middleware configuration
│ └── requirements.txt # Python backend dependencies
├── frontend/
│ ├── src/
│ │ ├── api/ # Axios API client functions
│ │ ├── components/ # UI components (goals dropdown, modals, cards)
│ │ ├── pages/ # Route views (Dashboard, Roadmap, Skills, Coach, etc.)
│ │ ├── store/ # Zustand state store with localStorage persistence
│ │ └── types/ # TypeScript interfaces and data types
│ ├── package.json # Node.js dependencies and scripts
│ └── vite.config.ts # Vite bundler configuration
└── README.md # Project documentation
All sensitive API keys and secrets (e.g., LLM_API_KEY, JWT_SECRET_KEY, DATABASE_URL) must be stored in environment variables and never committed to version control.
Example environment configuration for backend:
LLM_PROVIDER=google
LLM_MODEL=gemini-2.0-flash
LLM_API_KEY=your_gemini_api_key_here
DATABASE_URL=sqlite:///./neuronpath.db
JWT_SECRET_KEY=your_jwt_secret_key_hereMIT License — free for educational and commercial use.