🌍 TouristBuddy – Your Personalized Travel Assistant

🚀 Inspiration

Traveling to new places is exciting, but finding the best places to visit, planning routes, estimating costs, and managing time can be overwhelming — especially in a new city or country. The idea for TouristBuddy was born from this problem. I wanted to build a tool that acts like a local travel planner in your pocket — combining the power of Google Maps APIs with AI recommendations.

🧠 What I Learned

Throughout the development of TouristBuddy, I explored:

How to integrate multiple Google Maps Platform APIs for dynamic and responsive travel experiences.
Using Google Gemini APIs to generate contextual, real-time travel recommendations.
Designing intuitive user flows for seamless mobile and desktop experience.
Rate limiting, geolocation accuracy, and optimizing API calls for speed and budget efficiency.

🛠 How I Built It

TouristBuddy is built as a responsive, mobile-first web application with the following stack:

🌐 Frontend: HTML, TailwindCSS, JavaScript
🗺 Google Maps Platform APIs:
- Maps JavaScript API
- Places API
- Geocoding API
- Directions API
- Distance Matrix API
⚙️ Backend/Logic:
- Google Cloud Functions for API orchestration
- Supabase for storing saved plans and analytics
- Google Gemini API for smart, AI-powered travel tips and contextual insights
📱 UX/UI:
- Animations and effects using Lottie + responsive SVGs for delight
💡 AI-Powered Features:
- Auto-route planner
- Budget & time estimators
- Nearby suggestions with personalized tips

💻 Features

🔍 Discover Top Tourist Attractions based on city
🧠 AI-powered suggestions based on interest (nature, food, history, etc.)
📍 Smart Route Planner using Maps APIs
💸 Budget Estimator (includes travel, stay, and food)
📅 Travel Day Planner with durations & best travel times
📦 Save, share & print your trip plan (coming soon)

⚔️ Challenges I Faced

Rate Limiting: Google APIs have strict quotas. I had to create a logic that optimized calls and minimized redundant usage.
Location Precision: Extracting user intent and matching that to the right location in Maps was tricky. Places API helped, but refining results was key.
AI Contextualization: Making the Google Gemini AI responses more specific to city + interest + travel time required prompt tuning and memory logic.
Design Scope: Keeping it minimal but useful was a balancing act. I kept iterating using user feedback and rapid prototyping via Bolt.new.