Hackathon Winner: Tackling Prediabetes with Smart Agents

Imagine an AI system that doesn’t just track your health data, but truly understands it. In our latest “One-Shot” episode, we sit down with Varun Pant, whose team won the grand prize at the AI Tinkerers Humans-in-the-Loop Hackathon in Seattle (link) for developing SmartNourish, a health monitoring system that combines continuous glucose monitoring with AI assistance. Welcome to the bleeding edge of personalized healthcare technology.

Watch Now →

This isn’t just another health app. The system tackles a massive problem: over 1 in 3 Americans are prediabetic, facing potential long-term health complications. Traditional annual check-ups provide a snapshot, but SmartNourish offers continuous, personalized insights. By leveraging continuous glucose monitoring (CGM), genetic data, and advanced AI models like Anthropic’s Claude, Varun and team created an intelligent system that predicts, advises, and adapts.

Key Technical Highlights:

• Powered by LangChain and LangGraph
• Uses Anthropic’s Claude models for intelligent analysis
• Implements a sophisticated human-in-the-loop agent architecture
• Integrates real-time health data with predictive recommendations
• Agent design, tool integration, and human-in-the-loop strategies