EchoSphere – AI Voice Sales Agent for Car Dealerships
Link to open source: https://github.com/simrangupta8920-spec/HyundaiAI
Link to Live Project: https://hyundai-ai-flame.vercel.app/
EchoSphere is a real-time AI voice sales agent built for car dealerships, designed to solve a simple but costly problem: not every walk-in customer gets attention when staff are busy, leading to lost leads and inconsistent pitches.
With EchoSphere, a customer scans a QR code to start an instant, natural voice conversation with an AI assistant — no app download needed. The AI understands their budget and preferences, pulls live inventory data to recommend the best-matching car alongside a competitor comparison, and can offer pre-approved discounts within dealership-set rules. As the conversation progresses, it scores buyer intent in real time, and once a customer is ready to buy, automatically hands off a full summary to a human salesperson — so every lead is captured and every closer walks in prepared.
Built using Agora's Conversational AI voice engine, a FastAPI backend, PostgreSQL/Supabase for data, and a React + Tailwind frontend, with a retrieval-based (RAG) layer ensuring every recommendation is grounded in real, current inventory.
Built by Team Breeze the Builders for Agora Hackathon 2026.
This build was uploaded as a hackathon project
