Call Tour AI

Inspiration

Travel planning today is fragmented and time-consuming. Users often switch between multiple tabs for flights, hotels, and activities, making the process inefficient. We wanted to simplify this experience into a single, intelligent conversation using voice, making travel planning faster and more intuitive.

What it does

Call Tour AI is a real-time voice-based travel assistant that allows users to plan their entire journey through a single conversation. It understands user intent, captures preferences like budget and dates, and provides personalized recommendations for flights, hotels, cars, and activities instantly.

How we built it

We built a browser-based voice interface using HTML, CSS, and JavaScript. The backend is powered by FastAPI, handling APIs and system logic. We integrated Google Gemini for AI-driven intent understanding and response generation, along with LangGraph for workflow orchestration and LangChain for retrieval-based responses. LiveKit enables real-time voice streaming, while PostgreSQL stores conversation memory and ChromaDB supports semantic search.

Challenges we ran into

One of the biggest challenges was achieving low-latency real-time voice interaction. Synchronizing speech-to-text, AI processing, and text-to-speech without delays required careful optimization. Managing context across conversations and ensuring accurate intent detection was also challenging.

Accomplishments that we're proud of

We successfully built a working prototype that can handle real-time voice conversations for travel planning. The system provides instant recommendations and maintains conversational context, delivering a seamless user experience.

What we learned

We learned how to integrate multiple AI components into a real-time system, optimize latency, and design user-centric conversational workflows. We also gained experience in building scalable AI-powered applications.

What's next for Call Tour AI

We plan to enhance personalization, integrate live booking APIs, and expand support for multilingual voice interactions. In the future, we aim to scale this solution for enterprise travel agencies and real-world deployment.

Built With

• chromadb
• cloud
• css
• fastapi
• google-gemini
• html
• javascript
• langchain
• langgraph
• livekit
• postgresql
• python
• speech-to-text-(asr)
• text-to-speech-(tts)