🚨 "Design and development of an AI-powered learning path generator, Vocational Pathway Navigator with Dynamic Career Intelligence and NSQF-Integrated Learning Ecosystem"

🔹 Hackathon Origins

ShikshaDisha originally started as a hackathon project for SIH 2025 (Smart India Hackathon) by the DevBandits team. It was also built for the SUDHEE CBIT Hackathon 2026. Over time, it has evolved into a robust, AI-powered career navigation and labor market matching platform.

🏆 Built for SIH 2025 (Smart India Hackathon) as a Prototype

💡 Proposed Solution

✨ Key Features

• Smart Pathway Engine – AI analyzes learner profiles to generate personalized NSQF-aligned career routes
• AI Matching Engine – Smart course/curriculum matching based on user's input
• Career Journey Gamification – Achievement unlocks, skill mastery levels, industry challenges & leaderboards
• AI Learning Companion – Real-time guidance, industry alerts, skill forecasts & content recommendations

🎯 Problem Resolution

• Personalized NSQF Pathways – AI matches 50+ learner parameters to 139+ government courses with 95% accuracy
• Real-Time Market Alignment – Dynamic integration with labor market intelligence ensures pathway recommendations adapt to industry demands, emerging skills, and regional employment opportunities
• Multilingual Accessibility – 12+ Indian languages with voice navigation for diverse demographics

🔥 Unique Value Propositions

• Predictive Career Intelligence – AI forecasts employment probability & salary potential with 3-5 year projections
• Adaptive Pathway Evolution – Routes auto-adjust based on progress, industry changes & skill demands
• Gamified Engagement – Duels and streaks
• Cross-Sector Mobility – AI identifies transferable skills enabling seamless career transitions

📊 Feasibility and Viability

✅ Why It Works

• High Demand – Diverse learner backgrounds demand tailored skilling pathways
• Industry Alignment – Labour market intelligence ensures relevance to evolving job roles
• Future-Proofing – Adaptive AI pathways enable lifelong learning and stackable skills
• Institutional Backing – NCVET & MSDE integration provides credibility and adoption push

⚠️ Current Challenges & Risks

• User Trust: Learners may hesitate to rely on AI-driven career guidance
• Data Accuracy: Incomplete or outdated learner and labour market data may reduce recommendation quality
• Bias & Fairness: Risk of unequal opportunities if algorithms favor certain demographics or regions
• Long-Term Adoption: Sustaining engagement as career needs evolve requires continuous system updates

🛡️ Strategies to Overcome

• Trust Building: Explainable AI, counselor support, and transparent recommendation logic
• Data Quality: Regular updates from NSQF, labour market intelligence, and verified providers
• Fairness & Equity: Bias audits, inclusive design, and multilingual accessibility
• Sustained Engagement: Adaptive pathways, career milestone tracking, and continuous upskilling prompts

📚 Research & References

Key Supporting Market Facts

• 75% of Indian learners gain career benefits from AI-driven personalized paths
• 90% of employers prioritize NSQF-aligned micro-credentials in hiring
• India's EdTech market expected to surpass $10B by 2025, led by mobile-first apps
• AI-led adaptive learning speeds up skill acquisition by 30-40% versus traditional means
• 1+ billion Indian workers need reskilling by 2030 due to tech change and automation

Research Validation

• AI Personalization: AI creates customized learning paths that boost engagement and results
  [Frontiers in Education 2024 | Emerald AI in Education]
• Vocational Skilling & NSQF: NSQF links vocational training with industry and job market demands
  [ICRIER NSQF Note | IJFMR NSQF Implementation]
• Labour Market Intelligence: Real-time labour data integration is vital for future-ready skills
  [India Employer Forum 2025 | Economic Times EdTech]
• India EdTech Growth: EdTech is a $10B+ market growing quickly with AI-powered, mobile-first solutions
  [Market Research Future | HolonIQ Charts 2025]

💡 Takeaway: Research validates the importance of AI-personalized learning, NSQF compliance and scalable secure design for India's skill ecosystem.

📈 Success Metrics

• Pathway Recommendation Accuracy (Target: 95%)
• User Engagement Rate with AI Learning Companion
• NSQF Course Completion Rates
• Employment Outcome Tracking (6-month post-completion)
• Multi-language Adoption Metrics
• Labor Market Alignment Score
• User Satisfaction & Trust Scores

🎯 Impact on Target Audience

• Students & Youth: Personalized career pathways aligned with market demands
• Job Seekers: Data-driven career transitions with employment probability forecasts
• Working Professionals: Continuous upskilling with adaptive learning paths
• Educational Institutions: NSQF-integrated curriculum planning support
• Government Schemes: Enhanced effectiveness of Skill India missions through AI optimization

---

⚙️ Platforms

Platform	Supported?
Web (any browser with JS functionality) + Fully Responsive	✅
Android (non-natively through WebView)	✅

🔧 Development

🚀 Getting Started

Prerequisites

• Node.js 18+
• Python 3.11+
• Docker & Docker Compose
• PostgreSQL 15 (for local development)

Frontend

cd frontend-web
npm install
cp .env.template .env.local
npm run dev

Backend Services

Three microservices (backend_1-core_service, backend_2-ai_engine_service, backend_3-ai_companion_service) handle the split architecture for production. A monolithic backend/ service is also available and consolidates all functionality into one.

Option A: Monolithic Backend (Recommended for simplicity)

cd backend
python -m venv venv
venv\Scripts\activate  # Windows
# source venv/bin/activate  # Linux/Mac
pip install -r requirements.txt
uvicorn app.main:app --reload

Or with Docker:

docker compose up -d  # from project root — starts Postgres, Redis, backend, frontend

Option B: Split Microservices (Original)

See DEPLOYMENT_FULL.md for complete deployment instructions.

Production mode:

cd backend_1-core_service
docker-compose up -d
cd ../backend_2-ai_engine_service
docker-compose up -d
cd ../backend_3-ai_companion_service
docker-compose up -d

Development mode (with auto-reload on code changes):

cd backend_1-core_service
docker-compose --profile dev up -d
cd ../backend_2-ai_engine_service
docker-compose --profile dev up -d
cd ../backend_3-ai_companion_service
docker-compose --profile dev up -d

Docker Run (Standalone) — requires PostgreSQL, Redis already running locally:

cd backend_1-core_service
docker build -t shiksha-core .
docker run -d -p 8000:8000 --name shiksha-core \
  -e DATABASE_URL=postgresql://postgres:postgres@localhost:5432/shikshadisha \
  -e REDIS_URL=redis://localhost:6379/0 \
  shiksha-core

cd backend_2-ai_engine_service
docker build -t shiksha-ai-engine .
docker run -d -p 9000:9000 --name shiksha-ai-engine \
  -e REDIS_URL=redis://localhost:6379/0 \
  shiksha-ai-engine

cd backend_3-ai_companion_service
docker build -t shiksha-companion .
docker run -d -p 9001:9001 --name shiksha-companion \
  -e REDIS_URL=redis://localhost:6379/0 \
  shiksha-companion

Option C: Local Development (No Docker)

Requires PostgreSQL 15+ and Redis 7+ installed locally.

cd backend_1-core_service
python -m venv venv
venv\Scripts\activate
pip install -r requirements.txt
uvicorn app.main:app --reload

Hot Reload Development

Code changes are now mounted into containers via volumes. After making code changes:

# Rebuild and restart the service
docker-compose up -d --build

The --build flag rebuilds the image to pick up any dependency changes, while volumes mount your local code for live updates.

🏗️ Architecture

graph TB
    subgraph Clients
        Web[Web App<br/>Next.js]
    end

    subgraph "ShikshaDisha Backend Services"
        
        subgraph "backend-core :8000"
            API[Core API<br/>FastAPI]
            DB[(PostgreSQL)]
            Redis[(Redis)]
            Celery[Celery Workers]
            WS[WebSocket<br/>Real-time]
        end

        subgraph "backend-2-ai_engine_service :9000"
            Matcher[AI Matching<br/>Engine]
            FAISS[FAISS Index]
            Embed[Sentence<br/>Transformers]
            Behavior[Behavior<br/>Analyzer]
        end

        subgraph "/backend_3-ai_companion_service :9001"
            Chat[AI Companion<br/>Chat]
            Forecast[Skill<br/>Forecaster]
            Alerts[Industry<br/>Alerts]
            Rec[Content<br/>Recommender]
        end
    end

    Web --> API
    Web --> Matcher
    Web --> Chat
    
    API --> DB
    API --> Redis
    API --> Celery
    API --> WS
    
    Matcher --> FAISS
    Matcher --> Embed
    Matcher --> Behavior
    
    Chat --> Forecast
    Chat --> Alerts
    Chat --> Rec

Loading

Service Overview

Service	Port	Technology	Purpose
backend-core	8000	FastAPI + PostgreSQL	User management, actions, notifications, sessions, streaks
backend-2-ai_engine_service	9000	FastAPI + FAISS	Course matching, behavior analysis, recommendations
/backend_3-ai_companion_service	9001	FastAPI + Redis	AI chat, skill forecasting, alerts

Data Flow

1. User Actions → Core API → PostgreSQL + Celery Workers
2. Course Matching → AI Engine → FAISS Vector Search → Semantic Similarity
3. AI Companion → Skill Forecasts + Content Recommendations

Tech Stack

• Frontend: React, Next.js 14, TypeScript, TailwindCSS, shadcn/ui
• Backend: Python FastAPI (3 microservices)
• Database: PostgreSQL 15
• Cache/Queue: Redis 7
• AI/ML: Sentence Transformers, FAISS, scikit-learn
• Task Queue: Celery
• Container: Docker, GitHub Container Registry

📱 Screenshots *

Landing Page

👥 Our Team (DevBandits)

#	Team Member	Role	GitHub Profile
1	Fareed Ahmed Owais	🎯 Team Lead	🔗 FareedAhmedOwais
2	Abdur Rahman Qasim	🔎 Research Engineer	🔗 Abdur-rahman-01
3	Mohammed Saad Uddin	🚀 Full-stack + AI/ML Developer	🔗 saad2134

📊 Repo Stats

⭐ Star History

✨ Icon

🔰 Banner

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

• ✅ Commercial use
• ✅ Modification
• ✅ Distribution
• ✅ Private use
• ❌ Liability
• ❌ Warranty

---

✍️ Endnote

Developed with 💖 for the SIH 2025 (Smart India Hackathon) and SUDHEE CBIT Hackathon 2026, with heartfelt thanks for the opportunity to build and innovate.

---

🏷 Tags

#WebApp #SmartEducation #AIinEducation #PersonalizedLearning #SkillPathways #CareerGuidance #NSQFIntegration #VocationalEducation #AIPathGenerator #DigitalLearning #AdaptiveLearning #GamifiedLearning #TokenEconomy #AIMatching #SkillNavigator #FutureSkills #EdTechIndia #SkillForecasting #CareerIntelligence #MultilingualAI #SIH2025