ChatCX

ChatCX is an AI-powered agent designed to provide real-time insights and intelligence from Web3 Twitter. It tracks, analyzes, and delivers actionable data on emerging trends, narratives, and market-shifting conversations in the crypto space. The agent is built to help users stay ahead of the curve by providing pure, actionable insights powered by advanced RAG (Retrieval-Augmented Generation) and Reinforcement Learning algorithms.

---

🔗 Live API Base URL: https://autonome.alt.technology/chatcx-itferc/api

📢 Start using ChatCX: @ChatCX_Bot

📢 Start using Crypto Parrot: @CryptoParrotAI_Bot

📜 Docs: ChatCX Docs)

---

Features

• Real-time Data Tracking: Continuously monitors and analyzes posts from top Web3 influencers, analysts, and builders.
• AI-Powered Intelligence: Uses RAG and Reinforcement Learning to filter and extract actionable insights.
• Actionable Results: Provides instant, data-backed answers about trending tokens, narratives, and strategies.
• Multiple Products:
  • ChatCX Chat: Instant insights through a Telegram bot.
  • Crypto Parrot: AI-powered newsfeed delivering curated insights on Telegram.
  • API Access: Integrate ChatCX insights into your own agents, tools, and dashboards.

---

Tech Stack

• Frontend: React, TypeScript, TailwindCSS, Vite
• Backend: Node.js, Express, PostgreSQL
• AI/ML: OpenAI Embeddings, Hyperbolic API, RAG, Reinforcement Learning
• Telegram Bot: Telegraf, Telegram Web App
• Database: PostgreSQL
• DevOps: Docker, Docker Compose, Autonome by AltLayer (Hosting)
• User Management: Privy (Crypto Parrot)
• Other Tools: Puppeteer, PostHog, Sentiment Analysis, Keyword Extraction

---

Repository Structure

.
├── landing/ # Landing page for ChatCX
│   ├── src/ # React components and styles
│   ├── public/ # Static assets
│   ├── tailwind.config.js # TailwindCSS configuration
│   ├── vite.config.ts # Vite configuration
│   └── package.json # Frontend dependencies
│
├── crypto-parrot/ # Crypto Parrot Telegram bot and login
│   ├── server/ # Backend server for Crypto Parrot
│   ├── login/ # Login and authentication flow
│   └── package.json # Backend dependencies
│
├── bot/ # ChatCX Telegram bot
│   ├── src/ # Bot logic and services
│   └── package.json # Bot dependencies
│
├── service/ # Core AI service and Twitter pipeline
│   ├── src/ # AI logic, Twitter scraping, and API routes
│   ├── Dockerfile # Docker configuration for the service
│   ├── docker-compose.yml # Docker Compose setup
│   └── package.json # Service dependencies
|
├── docs/ # Docs for everything
│
└── README.md
│
└── LICENSE.md

---

Getting Started

Prerequisites

• Node.js (v18+)
• PostgreSQL
• Docker (optional) (node:18-bookworm-slim: strict requirement)
• Telegram Bot Token
• OpenAI API Key
• Hyperbolic API Key
• Privy API Key (for Crypto Parrot user management)

Installation

1. Clone the repository:

   git clone https://github.com/wei_b0/chatcx.git
   cd chatcx

2. Install dependencies:

   cd landing && npm install
   cd ../crypto-parrot/server && npm install
   cd ../../bot && npm install
   cd ../service && npm install

3. Set up environment variables:

   Create .env files in each directory (landing, crypto-parrot/server, bot, service) with the required environment variables (e.g., BOT_TOKEN, OPENAI_API_KEY, DATABASE_URL, PRIVY_API_KEY).

4. Run the services:

   • Landing Page:

     cd landing && npm run dev

   • Crypto Parrot Server:

     cd crypto-parrot/server && npm run dev

   • ChatCX Bot:

     cd bot && npm start

   • AI Service:

     cd service && docker-compose up --build

---

Technical Outline

1. Crypto Parrot

Server:

• Built with Express and TypeScript.
• Handles user registration, authentication, and rate limiting.
• Integrates with Privy for user management and authentication.
• Uses PostgreSQL for storing user data and rate limits.

Login:

• React-based login flow for Telegram users.
• Uses Telegram Web App SDK for seamless integration.
• TailwindCSS for styling.

2. ChatCX Telegram Bot

• Built with Telegraf and TypeScript.
• Provides real-time insights through a Telegram bot.
• Features:
  • /start command to initialize user and provide instructions.
  • Text-based queries to fetch AI-generated insights.
  • Rate limiting to manage free queries.
  • Error handling and user feedback.

3. AI Service

Twitter Pipeline:

• Scrapes Twitter for Web3-related tweets using agent-twitter-client.
• Extracts keywords, sentiment, and engagement metrics.
• Stores tweet data in PostgreSQL with vector embeddings.

AI Logic:

• Uses OpenAI embeddings for semantic search.
• Hyperbolic API for generating AI responses.
• RAG (Retrieval-Augmented Generation) and RL (Reinforcement Learning) for combining retrieval and generation.

API Routes:

• /chat: Handles user queries and returns AI-generated insights.
• /admin: Admin routes for managing API keys, updating datasets, and adding user credits.
• /health: Health check endpoint.

5. Database

• PostgreSQL is used for storing:
  • User data (Telegram IDs, credits, rate limits).
  • Tweet embeddings and metadata.
  • API keys and admin-related data.

6. DevOps

• Autonome by AltLayer:

  • The AI service and backend are hosted on Autonome, a decentralized hosting platform by AltLayer.
  • Autonome provides scalable, secure, and decentralized infrastructure for Web3 applications.

• Docker and Docker Compose for containerization and orchestration.

---

Contributing

Contributions are welcome! Please open an issue or submit a pull request for any improvements or bug fixes.

---

License

This project is licensed under the MIT License. See the LICENSE file for details.

---