AgentBot - Intelligent Energy Grid Analysis Chatbot

An advanced AI-powered chatbot system. that provides real-time energy grid analysis, weather information etc through an intuitive conversational interface.

Features

• Real-time Energy Grid Analysis - Monitor power demand, voltage readings, and grid status
• Weather Information - Get current weather data for any location worldwide
• Tool-Aware AI - Intelligent tool selection based on user queries
• Streaming Responses - Real-time token streaming for enhanced user experience
• Memory Management - Contextual conversations with session-based memory

Architecture

taskjs/
├── client/              # Next.js React frontend
├── server/              # Node.js Express backend
└── api_server/          # Data source API server

Frontend (Next.js + React)

• Modern React 18 with Next.js 14
• Tailwind CSS for responsive design
• Real-time chat interface with streaming support
• TypeScript for type safety

Backend Options

• Node.js Backend: Express.js with LangChain integration
• Data API: Separate data service for grid information

AI Integration

• LangChain Framework for AI orchestration
• Ollama Integration for local LLM deployment
• Tool Calling with function-based capabilities
• Memory Management for contextual conversations

Prerequisites

• Node.js 18+ and npm
• Python 3.10+ (for Python backend as chromadb is used by pip)
• Ollama installed and running
• OpenWeather API key (optional, for weather features)

Ollama Setup

1. Install Ollama from ollama.ai
2. Pull the required model:

   ollama pull llama3-groq-tool-use:8b
   # or
   ollama pull llama3.1:latest

3. Ensure Ollama is running on port 11434

Quick Start

1. Clone the Repository

git clone https://github.com/nafeu-khan/AgentBot.git
cd agentbot

2. Setup Environment Variables

Create .env files in the server/ directory:

3. Install Dependencies

Frontend Setup

cd client
npm install

Backend Setup (Node.js)

cd server
npm install

4. Start the Services

Option A: Node.js Backend

# Terminal 1: Start the frontend
cd client
npm run dev

# Terminal 2: Start the Node.js backend
cd server
npm run dev

# Terminal 3: Start the data API (optional)
cd api_server
npm run dev

5. Access the Application

• Frontend: http://localhost:3000
• Backend API: http://localhost:5001
• Data API: http://localhost:3001

Available Tools

The AI assistant has access to several specialized tools:

1. Grid Data Tool (getGridData)

• Retrieves real-time energy grid metrics
• Power demand, voltage readings, grid status
• Usage: "What's the current power demand?"

2. Weather Tool (getWeather)

• Fetches current weather for specified locations
• Requires specific location in the query
• Usage: "What's the weather in London?"

3. Company Info Tool (getCompanyInfo)

• Provides information
• Company details, services, contact information
• Usage: "Tell me about your company"

Model Configuration

The system supports various Ollama models:

• llama3-groq-tool-use:8b (recommended for tool calling)
• llama3.1:latest
• Any Ollama-compatible model with tool calling support

API Endpoints

Chat API

• POST /api/chat/ - Send chat message (streaming response)
• GET /health - Health check endpoint

Observability API

• GET /api/observability/memory/:sessionId - Get conversation memory
• GET /api/observability/system - System metrics

Usage Examples

Energy Grid Queries

User: "What's the current power demand?"
AI: [Calls getGridData tool and provides real-time metrics]

User: "Show me the latest voltage readings"
AI: [Retrieves and displays voltage data from grid]

Weather Queries

User: "What's the weather in Tokyo?"
AI: [Calls getWeather tool with location "Tokyo"]

User: "How's the weather?"
AI: "I need to know which city or location you'd like the weather for."

Company Information

User: "Tell me about your company"
AI: [Calls getCompanyInfo tool and provides company details]

Troubleshooting

Common Issues

1. CORS Errors

   • Ensure FRONTEND_URL matches your client URL
   • Check that both frontend and backend are running

2. Ollama Connection Issues

   • Verify Ollama is running: ollama serve
   • Check the OLLAMA_BASE_URL configuration

3. Tool Calling Issues

   • Ensure the model supports tool calling
   • Check that all required environment variables are set( see env.example)

4. Streaming Not Working

   • Check browser network tab for SSE connections