Inspiration

S.A.N.G.A.M is a comprehensive digital platform designed to enhance the spiritual journey of pilgrims during the Ujjain Mahakumbh 2028. During a family visit to Rishikesh last year, one of our elderly relatives got separated in the crowd. With poor network connectivity and unclear guidance, finding them took a long time and caused real panic. That experience stayed with us and inspired S.A.N.G.A.M - a platform meant to help families stay informed, connected, and safe during important spiritual journeys.

What it does

S.A.N.G.A.M serves as a complete digital companion for the Ujjain Mahakumbh, offering spiritual guidance, navigation assistance, community service opportunities, and safety features. The platform integrates traditional spiritual practices with modern technology to create an authentic and enriching pilgrimage experience.

📱 Applications

1. sangam_app - Flutter Mobile Application

  • Platform: Cross-platform mobile app (Android, iOS, Linux)
  • Technology: Flutter with Firebase backend
  • Purpose: Primary mobile experience for pilgrims

2. sangam_web - Flask Web Application

  • Platform: Web-based application
  • Technology: Flask with Firebase integration
  • Purpose: Web access and administrative features

✨ Key Features

🏛️ Sanskrti Module - Spiritual Heritage

  • Sacred Landmarks: Detailed information about 6+ major spiritual sites
  • Interactive Stories: Rich narratives about each location's spiritual significance
  • Video Content: YouTube integration for virtual tours
  • Punya Tracking: Earn spiritual merit points for visits
  • Visit Tracking: Mark and track visited sacred places

🤝 Seva Module - Community Service

  • Badge System: Earn service badges for completed seva
  • Progress Tracking: Monitor community service contributions
  • Real-time Updates: Live seva opportunities and locations

🏠 Home Module

  • Mahakumbh 2028 Schedule: Complete event calendar with 9 major events
  • User Dashboard: Personal punya points, badges, and progress
  • Quick Access: Navigation to all major features

📖 Sacred Streams Story Module

  • Interactive Storytelling: Animated story about Ujjain Mahakumbh
  • GSAP Animations: Scroll-based interactive animations
  • Visual Journey: Immersive storytelling experience
  • Educational Content: Learn about Ujjain's spiritual significance

👤 Profile Module

  • User Management: Complete profile with SANGAM ID
  • Achievement System: Display earned badges and punya points
  • Visit History: Track of visited landmarks and completed seva
  • Personal Stats: Comprehensive spiritual journey metrics

🤖 AI Chatbot - Sangam Sathi

  • Spiritual Guidance: AI-powered RAG based assistant for pilgrimage queries
  • Temple Information: Detailed information about Ujjain temples
  • Travel Tips: Practical advice for pilgrims
  • 24/7 Support: Always available digital companion

🆘 Safety Features

  • SOS Alert System: Emergency assistance with location sharing
  • Family Finder: Locate family members in crowds
  • Crowd Navigation: Real-time guidance for managing large gatherings
  • Emergency Contacts: Quick access to help services

How we built it

Instead of writing everything line-by-line, we followed these principles:

  • Start with high-level intent, not syntax
  • Describe what we want, not how to implement it
  • Iterate fast using prompt → generate → tweak → commit
  • Optimize for momentum and creativity, especially under hackathon constraints

This allowed the architecture to emerge organically rather than being rigidly planned upfront.

🤖 AI Models & Tools Used

1️⃣ GitHub Copilot — Our Primary Vibe-Coding Engine

  • Used continuously as an AI pair programmer
  • Powered by *OpenAI GPT-5.1 Codex *
  • Helped us:

    • Scaffold Flutter UI screens from comments
    • Generate Flask backend routes from natural language
    • Write Firebase authentication & data logic
    • Refactor and clean hackathon-speed code
  • Claude Sonnet 4.5 / Claude Haiku 4.5
    Used for cleaner reasoning, rewriting complex logic, and improving readability of AI-generated code.

  • Gemini 2.5 Pro / Gemini Flash
    Used for fast experimentation, structured responses, and alternative reasoning paths during development.

  • Prompt-in-Code Workflow
    High-level comments and intent-based prompts were written directly in code files and expanded by AI.

Challenges we ran into

  • Aligning AI-generated code with hackathon-speed changes
  • Maintaining consistency across AI outputs from different models
  • Debugging edge cases in rapidly generated backend logic
  • Prompt tuning to get desired UI and chatbot behavior
  • Integrating AI-generated components smoothly end-to-end

Accomplishments that we're proud of

  • Designed and built a complete digital ecosystem for Mahakumbh 2028, combining spiritual guidance, navigation, safety, and community engagement in a single platform
  • Developed cross-platform mobile and web applications (Flutter + Flask) with seamless Firebase integration for authentication, data, and storage
  • Implemented a RAG-based AI chatbot (Sangam Sathi) that provides spiritual guidance, temple information, travel tips, and 24/7 support
  • Created feature-rich modules like Sanskrti (spiritual landmarks & punya tracking), Seva (community service & badges), and interactive storytelling with animations
  • Integrated safety-first features including SOS alerts, family finder, and crowd navigation concepts for large gatherings
  • Successfully delivered an end-to-end, scalable, and user-ready product using a vibe-coding approach within hackathon constraints

What we learned

  • How to use vibe coding effectively by treating AI as a real-time collaborator
  • Building and shipping a full-stack product quickly under hackathon constraints
  • Designing AI-powered features (RAG chatbots) that are practical and user-focused
  • Integrating multiple modules (spiritual, safety, community) into one cohesive system
  • Balancing speed, creativity, and code quality in rapid development cycles

What's next for SANGAM

  • Enhance real-time navigation and crowd management using live location data
  • Expand AI capabilities with multilingual support and smarter contextual guidance
  • Add offline-first features for low-connectivity scenarios during Mahakumbh
  • Integrate official authorities and seva organizations for verified updates
  • Scale the platform for large user volumes closer to Mahakumbh 2028

Built With

  • css
  • dart
  • firebase-authentication
  • firebase-storage
  • firestore
  • flask
  • flutter
  • github-copilot
  • gsap
  • html
  • n8n
  • openai-gpt-class-models
  • python
  • retrieval-augmented-generation-(rag)
Share this project:

Updates

Submission history