Inspiration
As avid travelers, we've experienced the frustration of juggling multiple apps, websites, and spreadsheets to plan a single trip. We wanted to create a unified platform that combines the power of modern CMS, lightning-fast search, and AI to make travel planning effortless and enjoyable. The challenge was clear: how can we help travelers discover amazing destinations and plan perfect trips in minutes, not hours?
What it does
AtlasGo is an intelligent travel discovery platform that revolutionizes how people explore and plan trips:
🔍 Lightning-Fast Search - Find places across 48+ destinations in 6 countries with sub-50ms response times powered by Algolia. Search by location, type, price, interests, and even find places near you with geo-search. 📝 Smart Content Management - 70+ content items including places, travel guides, local tips, food guides, events, and neighborhood profiles—all managed seamlessly through Storyblok CMS. 🤖 AI-Powered Trip Planning - Generate personalized day-by-day itineraries using GROQ AI (Llama 3.3). Tell us your preferences, dates, and interests, and get a complete trip plan with real places from our database. 🗺️ Interactive Maps - Visualize destinations with Leaflet maps, cluster markers, and detailed place information. ⭐ Curated Collections - Discover handpicked destinations through collections like "World-Class Museums," "Natural Wonders," "Hidden Gems," and more.
How we built it
Tech Stack:
- Frontend: Next.js 14 with TypeScript and TailwindCSS for a modern, responsive UI
- CMS: Storyblok for flexible, structured content management with multiple content types
- Search: Algolia for instant search with geo-queries, faceted filters, and typo-tolerance
- AI: GROQ AI (Llama 3.3 70B) for intelligent itinerary generation
- Maps: Leaflet with OpenStreetMap for interactive visualizations
- Deployment: Railway for seamless hosting
Architecture:
- Content Layer - Storyblok stores all places, guides, tips, and collections with rich metadata
- Search Layer - Custom indexing pipeline syncs Storyblok content to Algolia with geo-coordinates
- AI Layer - GROQ AI analyzes user preferences and generates contextual itineraries using real place data
- Presentation Layer - Next.js App Router with server and client components for optimal performance
Challenges we ran into
- Content Relationships - Managing relationships between places, collections, and guides in Storyblok required careful schema design. We solved this by using slug-based references and dynamic content fetching.
- Geo-Search Accuracy - Ensuring accurate GPS coordinates for all 48 places across 6 countries was time-consuming. We manually verified each location and implemented fallback mechanisms.
- AI Context Management - Getting GROQ AI to generate realistic itineraries with actual places required careful prompt engineering and context management. We created a structured prompt system that includes place metadata.
- Search Performance - Balancing rich content with fast search required optimizing Algolia index settings, implementing custom ranking, and fine-tuning relevance parameters.
- Map Clustering - Implementing Leaflet marker clustering with custom popups and handling dynamic updates required solving several initialization timing issues.
Accomplishments that we're proud of
✨ 48+ Places Across 6 Countries - Built a comprehensive global database with accurate data, GPS coordinates, and high-quality images ⚡ Sub-50ms Search - Achieved lightning-fast search performance with complex filtering and geo-queries 🎨 Professional UI/UX - Created a modern, accessible interface with custom SVG icons, smooth animations, and responsive design 🤖 Working AI Integration - Successfully integrated GROQ AI to generate realistic, personalized itineraries using real place data 📚 Rich Content Ecosystem - Implemented 8 different content types beyond just places: travel guides, local tips, food guides, events, neighborhoods, collections, testimonials, and FAQs 🔄 Automated Pipeline - Built scripts for content management, reindexing, and data synchronization 🌐 Scalable Architecture - Designed a system that can easily expand to more countries and content types
What we learned
- How to architect a multi-layered content system with Storyblok's flexible CMS
- Advanced Algolia search techniques including geo-queries, custom ranking, and faceted navigation
- Next.js 14 App Router patterns and optimization strategies
- Real-time data synchronization between CMS and search infrastructure
What's next for AtlasGo
🎯 Advanced Personalization - Machine learning for better recommendations 🏨 Booking Integration - Partner with hotels, tours, and restaurants 🌐 Multi-language Support - Translate content into 10+ languages 📸 User-Generated Content - Photo uploads, trip reports, tips 🤝 Social Features - Follow travelers, share trips, collaborative planning 🚀 AR Navigation - Augmented reality for exploring destinations 🎮 Gamification - Badges, achievements, travel challenges 🌱 Sustainable Travel - Carbon footprint tracking, eco-friendly options 🤖 Voice Assistant - "Hey AtlasGo, plan a 5-day trip to Japan" 🌐 Global Community - Connect travelers worldwide
🏆 Why AtlasGo Stands Out For the Hackathon:
- Demonstrates Storyblok's flexibility with 8 content types
- Showcases Algolia's power with geo-search and facets
- Real-world application solving actual user problems
- Production-ready architecture and design
- Scalable foundation for future growth
Built With
- algolia's
- groq
- html5/css3
- javascript
- leaflet.js
- leaflet.markercluster
- next.js
- openstreetmap
- react
- storyblok
- tailwindcss
- typescript


Log in or sign up for Devpost to join the conversation.