Hubli and Mysore both have incredibly rich culinary identities, but tourists visiting Hubli often struggle to find authentic, locally-prepared Mysore specialties like the perfect Mysore Masala Dosa or genuine Mysore Pak. We wanted to build a bridge between these two food cultures. Mysore Food AI was inspired by the idea of using modern artificial intelligence to act as a hyper-local culinary guide, helping travelers experience authentic flavors without getting lost in generic restaurant recommendations.
⚙️ What it does
Mysore Food AI is an intelligent assistant designed specifically for tourists in Hubli. Users can chat naturally with an AI assistant to discover restaurants, hidden gems, and local eateries that serve authentic Mysore-style dishes. The platform features:
- Natural Language Recommendations: Ask for specific cravings (e.g., "Where can I get a crispy dosa with authentic red chutney nearby?").
- Smart Mapping Integration: Visually plots the best culinary routes and restaurant pins directly across Hubli.
- Cultural Context: Provides historical background or fun facts about the dishes you are exploring.
🛠️ How we built it
The core backend logic relies on a smart AI orchestration layer using Python. We utilized specialized semantic search to index restaurant menus and local food reviews across Hubli. The interface is clean and user-friendly, designed to give tourists a seamless mobile experience with interactive map integration to handle real-time location pinning.
🚧 Challenges we ran into
One of our biggest hurdles was data accuracy—ensuring the AI didn't just recommend any South Indian restaurant, but specifically filtered for authentic Mysore preparation styles. Debugging the logic to cross-reference user location with real-time restaurant availability took a lot of fine-tuning, but we solved it by setting strict filtering parameters within our data retrieval pipeline.
🎉 Accomplishments that we're proud of
We successfully built a highly niche, culturally aware AI assistant that feels like talking to a local food expert. Getting the mapping system to smoothly display custom pins alongside the AI conversational interface was a massive win for our team's workflow.
📚 What we learned
We gained deep experience in bridging conversational AI models with geographical data APIs. We also learned a lot about the nuances of local regional cuisines and how valuable hyper-localized data can be for travelers.
🔮 What's next for Mysore Food AI
We plan to expand the platform to support more regional culinary crossover guides across India. We also want to implement real-time crowd-sourced reviews and allow local vendors to update their daily specialty menus directly through the AI.
