A full-stack AI chat platform powered by multiple LLMs — with RAG, real-time internet search, vision, document understanding, and more.
| Login | Register |
|---|---|
 |
 |
| Chat UI | Settings |
|---|---|
 |
 |
| About | Mobile View |
|---|---|
 |
 |
ZErio AI is a production-grade AI assistant platform that connects multiple state-of-the-art language models under one sleek interface. It goes beyond a simple chatbot — with RAG-powered document intelligence, real-time web search, image understanding, and live streaming responses — all secured with Redis-backed authentication and delivered with socket-powered real-time communication.
Switch between the world's most powerful AI models in one click:
| Model | Provider |
|---|---|
| Gemini 1.5 Pro / Flash | |
| GPT-4o / GPT-3.5 | OpenAI |
| Mistral Medium | Mistral AI |
| Command R | Cohere |
| DeepSeek V1 | DeepSeek |
Upload your files and chat with them directly. ZErio AI parses and understands:
- 📄 PDF documents
- 📝 DOCX / Word files
- 🗂️ JSON data files
Powered by Tavily Search API + LangChain — ZErio AI can search the web in real-time and ground responses in current, accurate information.
Upload any image and ask questions about it. ZErio AI uses multimodal LLMs to analyze, describe, and reason over visual content.
Responses stream token-by-token in real time via Socket.io — no waiting, just instant intelligence.
- JWT-based authentication
- Email verification flow with beautiful HTML emails
- Redis-powered token blacklisting on logout
- HttpOnly cookie sessions
All uploaded files and images are stored and managed via ImageKit CDN — fast, optimized, and globally delivered.
| Technology | Purpose |
|---|---|
| Node.js + Express | REST API server |
| MongoDB + Mongoose | Primary database |
| Redis | Session management & logout blacklist |
| Socket.io | Real-time LLM streaming |
| LangChain | LLM orchestration & RAG pipeline |
| Tavily Search API | Real-time internet search |
| ImageKit | File & image storage CDN |
| Brevo (SMTP) | Transactional email delivery |
| JWT | Authentication tokens |
| Technology | Purpose |
|---|---|
| React + Vite | UI framework |
| Tailwind CSS | Styling |
| Axios | HTTP client |
| React Router | Client-side routing |
| Socket.io Client | Real-time communication |
| Integration | Usage |
|---|---|
| Google Gemini | Primary LLM |
| OpenAI GPT | Alternative LLM |
| Mistral AI | Alternative LLM |
| Cohere Command | Alternative LLM |
| DeepSeek | Alternative LLM |
| Tavily | Web search grounding |
| LangChain | RAG + orchestration |
User Request
│
▼
React Frontend (Vite)
│ Socket.io + REST
▼
Express Backend
├── Auth (JWT + Redis + Brevo Email)
├── LLM Router (Gemini / GPT / Mistral / Cohere / DeepSeek)
├── RAG Pipeline (LangChain + Vector Store)
│ └── PDF / DOCX / JSON Parser
├── Web Search (Tavily API)
├── Vision (Multimodal LLM)
└── File Storage (ImageKit CDN)
│
▼
MongoDB (Chats, Users, Documents)
Redis (Session Blacklist)
- ✅ Production Deployed — Live on Render with real users
- ✅ Multi-LLM Architecture — Engineered to swap models without changing business logic
- ✅ RAG Implementation — Built a full retrieval-augmented generation pipeline from scratch
- ✅ Real-time Streaming — Socket.io for live token-by-token LLM output
- ✅ Secure by Design — Redis token blacklisting, HttpOnly cookies, email verification
- ✅ Full-Stack — End-to-end ownership from DB schema to UI animations
- ✅ Clean Code — Modular service-based backend, component-driven frontend
Harsh Patel
Full-Stack Developer · AI Engineer
Built for the future. ⚡
© 2026 ZErio AI — All rights reserved.