RAG-Based Intelligent Q&A & Content Platform

An AI platform that combines multi-source knowledge retrieval with LLM-powered content generation. Acts as an intelligent assistant capable of answering questions, retrieving accurate information from diverse data sources, and generating context-aware content.

Built with FastAPI for APIs, Streamlit for UI, and LangChain for managing advanced LLM workflows, featuring integrated RAG pipelines for document-aware interactions.

Key Technologies: FastAPI • Streamlit • LangChain • OpenAI GPT-3.5 • Ollama Llama2 • Vector Databases

Use Cases: Multi-source chatbots, document-based Q&A, enterprise knowledge assistants, AI-driven content generation, educational and research tools

Features

Feature	Implementation	Purpose
LLM-based Chatbot	FastAPI APIs + Streamlit UI	Conversational interface with contextual understanding
RAG Pipeline	Vector databases + LangChain	Accurate document-based Q&A with source context
Multi-source Agents	Wikipedia + JSON/Text/Web/PDF	Pulls information from multiple data sources
Content Generation	LLM APIs (OpenAI, Llama2)	Generates high-quality content and summaries
Scalable Architecture	FastAPI backend + modular design	Flexible for enterprise or product integration

Professional web interface for generating essays and poems with AI

Project Structure

├── api/
│   ├── app.py          # FastAPI server with essay/poem endpoints
│   └── client.py       # Streamlit web interface
├── agents/
│   └── agents.ipynb    # Multi-source RAG agents with Wikipedia integration
├── rag/
│   ├── simplerag.ipynb # Basic RAG pipeline with vector database
│   ├── chain.ipynb     # Advanced RAG chain implementations
│   └── *.pdf, *.json   # Sample documents and data sources
├── chatbot/
│   └── app.py          # Experimental chatbot interface
├── requirements.txt    # Python dependencies
└── .env               # API keys (create this file)

Key Components:

• API: Production-ready FastAPI endpoints with web interface
• Agents: Advanced AI agents with multi-source data retrieval
• RAG: Retrieval-Augmented Generation experiments and implementations
• Chatbot: Interactive conversational AI prototypes

Prerequisites

Software Requirements:

• Python 3.8+ (python --version)
• Git (optional, for cloning)

API Keys Required:

• OpenAI API Key - ~$0.002 per essay, $5 free credit
• LangChain API Key - Free for monitoring
• Ollama - Free local AI model

Quick Start

1. Clone and Install

   git clone <repository-url>
   cd <repository-directory>
   pip install -r requirements.txt

2. Setup Environment Create .env file:

   LANGCHAIN_API_KEY=your-langchain-key
   OPENAI_API_KEY=your-openai-key
   LANGCHAIN_PROJECT=langchain_project
   LANGCHAIN_TRACING_V2=true

3. Install Ollama (for poems)

   # Download from https://ollama.ai/
   ollama pull llama2

4. Run the Application

   # Terminal 1: Start API server
   uvicorn api.app:app --reload
   
   # Terminal 2: Start web interface
   streamlit run api/client.py

5. Access the App

   • Web Interface: http://localhost:8501
   • API Documentation: http://localhost:8000/docs

Usage

Web Interface

Visit http://localhost:8501 and use the intuitive forms to generate content.

API Endpoints

Essay Generation:

curl -X POST "http://localhost:8000/essay/invoke" \
  -H "Content-Type: application/json" \
  -d '{"topic": "Renewable Energy", "word_count": 150}'

Poem Generation:

curl -X POST "http://localhost:8000/poem/invoke" \
  -H "Content-Type: application/json" \
  -d '{"topic": "Autumn Leaves", "audience": "young adult"}'

Troubleshooting

Issue	Solution
API Key Error	Verify .env file format, check OpenAI credits, restart server
Ollama Not Found	Run ollama serve and ollama pull llama2
Port in Use	Use different ports: --port 8001 or --server.port 8502
Module Not Found	Run pip install -r requirements.txt

Debug Steps:

1. Check terminal logs for detailed errors
2. Verify you're in the correct directory
3. Test API endpoints at http://localhost:8000/docs

Development

Restart Checklist:

• Python installed (python --version)
• API keys valid (check OpenAI billing)
• Ollama running (ollama serve)
• Dependencies installed (pip install -r requirements.txt)

Extension Ideas:

• Additional AI models (Claude, Gemini)
• User authentication and content saving
• Multiple writing styles and formats
• Batch processing capabilities

License

MIT License - see LICENSE file for details.