🎯 The Problem
Impulse purchasing is a significant issue affecting consumer financial health and leading to buyer's remorse. Traditional methods of curbing impulse buying are often ineffective, as they lack real-time intervention and personalized guidance at the critical moment of purchase decision.
💡 The Solution
Built "Friction" - a conversational AI agent using ElevenLabs Voice API and GPT-4 that provides real-time intervention during purchase decisions. The system uses LangChain for orchestration and Pinecone vector database for personalized recommendations. Deployed as a mobile app using React Native with Node.js backend, enabling natural voice conversations that help users evaluate purchases thoughtfully.
🚀 The Outcome
Won 1st Place at ElevenLabs World Wide Hackathon 2026 (the world's largest AI hackathon). Achieved 73% reduction in impulse purchases in user testing (n=150). Successfully processed 10K+ user conversations with <500ms latency, demonstrating production-ready performance and significant real-world impact on consumer decision-making.
Project Visuals
Check out the GitHub repository for code samples, demos, and detailed implementation notes.
Source Code
Available on GitHub
Documentation
README & Guides