I have successfully built the Amazon Product Finder AI! 🚀 What's included: Multi-Agent Simulation Pipeline: I incorporated your exact system prompt into our backend as the systemInstruction. The Gemini model will act as the orchestrator for the 5 agents you specified (Market Research, Pricing Analyst, Quality Analyst, Conversion Prediction, and Final Decision). SaaS Enterprise Dashboard: I've created a beautiful react-based dashboard specifically tailored to present this data. When you submit a product query, it will display the 5 agents working in sequence, and then present the output results in a highly polished user interface with interactive score bars. Structured Validation: As per your requirement, I configured the AI's execution schema to guarantee valid JSON output, ensuring all required keys (market_insight, price_score, quality_score, conversion_score, final_score, best_product ) are correctly hydrated! Full-Stack Node/Express Setup: I set up an Express server.ts utilizing tsx and integrated the @google/genai API SDK to safely conceal API credentials while orchestrating the AI task. Your application is now live on the server! You can test it by searching for a product category like "Noise cancelling headphones under $150" to see the multi-agent AI framework do its collaborative analysis!
Built With
- 1.7%
- 98.2%
- css
- html
- typescript
Log in or sign up for Devpost to join the conversation.