Inspiration
Planning a trip is stressful — especially for students balancing budget, preferences, and time. We wanted to build a personal AI travel assistant that actually understands you — your pace, your interests, and your style.
What it does
wAI Travel Planner generates personalized travel itineraries using Amazon Bedrock and DynamoDB. Users can create a profile with their interests (like hiking or food), and the AI crafts a trip that matches their travel style and budget. It stores and adapts itineraries over time, making future trips smarter.
How we built it
We used a serverless AWS architecture:
AppSync (GraphQL API) for querying and mutations
Lambda for logic (itinerary generation, profile updates)
DynamoDB for user and itinerary storage
Bedrock (Claude 3.5 Sonnet) for generating itinerary content The frontend is a Next.js app connected via Apollo Client and Amplify Auth for Cognito-based login.
Challenges we ran into
Configuring IAM roles and connecting AppSync resolvers to Lambda securely was harder than expected. Bedrock setup also took fine-tuning to make the model context-aware of each user’s preferences.
Accomplishments that we're proud of
We built a fully working AI-powered travel planner with persistent user profiles and personalized trip generation — all serverless and scalable on AWS.
What we learned
We learned how to connect multiple AWS services (AppSync, Bedrock, DynamoDB, Lambda, Amplify) into a smooth pipeline, and how to manage permissions and environments efficiently.
What's next for wAI Travel Planner
We plan to add live data fetching through an MCP server — pulling real attractions, restaurants, and flights in real time — and support collaborative trip planning with friends.
Log in or sign up for Devpost to join the conversation.