E-Commerce AI Customer Support Assistant

E-Commerce AI Customer Support Assistant This project is a voice-enabled, multilingual AI customer support assistant designed specifically for e-commerce platforms. It provides instant, intelligent responses to user queries using a combination of: Issue Classification (e.g., Order Issues, Returns, Payments) Sentiment Analysis to gauge customer mood RAG-based FAQ Matching using sentence embeddings LLM-powered responses using Groq’s LLaMA 3 API Real-Time Web Search via SERP API and Tavily for unseen queries Voice-to-Text & Text-to-Speech (gTTS) support User Profile Memory: Tracks language preference, common queries, sentiment history, and issue types Optional Agent-Based Architecture using Fetch.ai’s uAgents framework for modular agent communication (e.g., FAQ agent) The assistant works in multiple languages (English, Hindi, Spanish, French), and replies can be translated based on user preference. It supports rich user interactions via voice input and audio replies. A persistent user profile is stored and visualized in the UI, allowing personalized, context-aware interactions. This solution is designed to scale with plug-and-play modules for travel, fintech, logistics, and more — making it a powerful foundation for real-world B2C applications.