Ramble

Response generated by FastAPI
Logs

Inspiration

India is a land of deep cultural, spiritual, and regional diversity. First-time travelers and foreigners often find it difficult to go beyond the surface-level tourist experience. We were inspired to build Ramble to offer these travelers a voice-first, deeply local experience that helps them explore not just famous spots, but also hidden gems, regional customs, food, and avoid scams — all through natural conversation.

What it does

Ramble is a voice-based regional travel assistant. Users select a region (e.g., Amritsar), and the app generates a region-specific homepage that highlights:

Sacred places
Must-visit spots with visuals, distance, cost, and brief info (in card format)
Hidden gems known only to locals
Food recommendations
Cultural and safety tips Users can ask questions via voice input in a chat-like interface, and get responses in voice and text, pulled from reliable sources like the Sonar API. When users select a location card, Ramble generates checkpoints within that spot—important areas to cover so nothing is missed. After completing a place, the chat reflects it as "completed". It also supports:
Live questions (e.g. “What’s the weather like today?”)
Local phrases with pronunciation help
Bargaining tips to avoid getting scammed
**Voice- & Chat-Based interactions for accessibility

How we built it

We built Ramble with a voice-first UX in mind, layering chatbot interaction over region-specific modules. The core elements include:

Voice input and output with natural, conversational responses
Dynamic homepages based on selected Indian cities or regions
Cards for attractions, with API-generated data like prices, maps, and highlights
A custom checkpoint system that breaks down each location into must-see parts
A culture and info section with local phrases, photo spots, food, and etiquette

Integration with Sonar API for intelligent, location-aware search responses

Challenges we ran

Designing a voice-first UI that doesn’t feel robotic or clunky
Condensing deep regional information into short, meaningful conversations
Ensuring Sonar API results stay region-relevant and personalized
Building trust by accurately guiding users through unknown, local cultural spots
Avoiding information overload while still covering all important areas

Accomplishments that we're proud of

Created a fully voice-based travel assistant tailored for India
Enabled checkpoint-based guidance to simplify on-ground navigation
Built a cultural layer into the product that includes phrases, traditions, and etiquette
Highlighted hidden cultural gems not typically found online, making the travel experience truly immersive

What we learned

Voice-first interactions make travel tools more inclusive and comfortable, especially for foreigners
Travelers appreciate knowing not just “where” to go, but how to behave, speak, and interact
Hidden gems are often more impactful than mainstream attractions, especially in culturally rich areas like Amritsar
Region-specific personalization increases user trust and engagement

What's next for

Expanding to more Indian cities and states with deeply localized content
Adding multilingual support (e.g., Hindi, Punjabi, Tamil) for both voice input and output
Partnering with local guides and historians for authentic hidden gem data
Improving map-based checkpoint tracking and offline voice support
Scaling user feedback integration to enhance real-time recommendations