FastAPI Chatbot

A RESTful API for a chatbot built with FastAPI, providing structured endpoints for chat interactions and content summarization.

Project Structure

.
├── app/
│   ├── api/
│   │   ├── __init__.py
│   │   ├── dependencies.py
│   │   ├── errors.py
│   │   └── routes/
│   │       ├── __init__.py
│   │       ├── chat.py
│   │       └── summarizer.py
│   ├── core/
│   │   ├── __init__.py
│   │   ├── config.py
│   │   ├── logger.py
│   │   └── errors.py
│   ├── models/
│   │   ├── __init__.py
│   │   └── schema.py
│   ├── services/
│   │   ├── __init__.py
│   │   ├── chatbot.py
│   │   └── summarizer.py
│   └── __init__.py
├── logs/
├── tests/
│   ├── __init__.py
│   └── test_api.py
├── .env
├── .gitignore
├── main.py
├── requirements.txt
└── README.md

Setup Instructions

Prerequisites

• Python 3.8+
• pip (Python package installer)
• Ollama (for LLM functionality)

Installing and Setting up Ollama

Ollama is required to run the local LLM used by the chatbot and summarization features.

Windows

# Download and install Ollama from the official website
# https://ollama.com/download/windows

# After installation, pull the required model
ollama pull llama3:latest

macOS

# Install Ollama using Homebrew
brew install ollama

# Or download from the official website
# https://ollama.com/download/mac

# Pull the required model
ollama pull llama3:latest

Linux

# Install Ollama
curl -fsSL https://ollama.com/install.sh | sh

# Pull the required model
ollama pull llama3:latest

After installing, make sure the Ollama service is running:

# Start Ollama service
ollama serve

The Ollama API should be available at http://localhost:11434

Setting Up a Virtual Environment

Windows

# Create a virtual environment
python -m venv venv

# Activate the virtual environment
venv\Scripts\activate

# If you're using PowerShell instead of Command Prompt
# venv\Scripts\Activate.ps1

macOS/Linux

# Create a virtual environment
python3 -m venv venv

# Activate the virtual environment
source venv/bin/activate

Installing Dependencies

Once your virtual environment is activated, install the required packages:

pip install -r requirements.txt

# Download SpaCy language model
python -m spacy download en_core_web_sm

Running the Application

Development Server

uvicorn main:app --reload

The API will be available at http://127.0.0.1:8000

API Documentation

FastAPI automatically generates API documentation using Swagger UI and ReDoc:

• Swagger UI: http://127.0.0.1:8000/docs
• ReDoc: http://127.0.0.1:8000/redoc

Environment Variables

Create a .env file in the root directory with the following variables:

APP_ENV=development
DEBUG=true
LOG_LEVEL=INFO

API Endpoints

Chat Endpoints

Endpoint	Method	Description
/api/v1/chat/message	POST	Send a message to the chatbot
/api/v1/health	GET	Check API health status

Summarization Endpoints

Endpoint	Method	Description
/api/v1/summarize/youtube	POST	Summarize a YouTube video from its URL
/api/v1/summarize/pdf	POST	Summarize a PDF document from an uploaded file
/api/v1/summarize/text	POST	Summarize plain text content

Browser-Based Session Management

The chatbot uses cookies to automatically track user sessions:

• No login or user accounts required
• Sessions are created and managed automatically via browser cookies
• Cookies expire after 90 days
• When a user visits for the first time, a unique session ID is generated
• All subsequent requests from the same browser reuse the same session
• No need to manually provide or track session IDs

This approach simplifies the user experience while maintaining conversation context across visits.

Summarization Features

The API provides content summarization with the following options:

Summary Length Options

• Short: 1-2 paragraphs
• Medium: 3-5 paragraphs
• Long: 5-7 paragraphs

Summary Style Options

• Narrative: Flowing text format
• Bullet Points: Bullet point format
• Academic: Formal, structured format
• Simplified: Easy to understand, non-technical language

YouTube Video Summarization

Extracts and summarizes transcripts from YouTube videos, with options to include timestamped sections. Features:

• Automatic transcript extraction and translation if needed
• Video metadata extraction
• Key points identification
• Timestamped section detection

PDF Document Summarization

Extracts and summarizes text from PDF documents, with options to specify page ranges.

Text Summarization

Summarizes plain text content with various style and length options.

Form-Based API

All endpoints use form-based input with the following benefits:

• Easy to test directly in the Swagger UI
• Dropdown selectors for summary length and style
• File upload for PDF documents
• Optional fields with clear documentation

Technical Implementation

• Text processing and summarization use SpaCy for natural language processing
• YouTube transcripts are extracted using the YouTube Transcript API
• PDF text extraction uses pdfplumber
• No authentication required - completely open API
• Cookie-based session tracking

Error Handling

This project includes comprehensive error handling with:

• Custom exception classes
• Proper HTTP status codes
• Structured error responses
• Request validation

Logging

Logs are stored in the logs/ directory with different levels (INFO, WARNING, ERROR) and automatic rotation.