Vsai2004/AI_Customer_Support_Bot

AI Customer Support Bot

An intelligent customer support bot that can answer questions based on your documents using advanced NLP models.

Features

• —Document Processing: Upload PDF or text files, or use built-in sample documents
• —Intelligent Q&A: Uses DistilBERT and sentence transformers for accurate responses
• —Feedback Loop: Automatically improves responses based on simulated feedback
• —Real-time Chat: Interactive chat interface with conversation history
• —Analytics: Track response times and conversation statistics

How to Use

1. 1.Choose a Document Source:
2. 2.Select from sample documents (FAQ, API docs, Product guide)
3. 3.Upload your own PDF or text file
4. 4.Paste text directly

1. 1.Ask Questions: Once the bot is initialized, type your questions in the chat interface

1. 1.View Results: Get instant AI-powered responses with context from your documents

Sample Questions to Try

• —"How do I reset my password?"
• —"What's the refund policy?"
• —"How do I contact support?"
• —"What are the API rate limits?"
• —"How do I get started?"

Technical Details

• —Embedding Model: sentence-transformers/all-MiniLM-L6-v2
• —QA Model: distilbert-base-uncased-distilled-squad
• —Vector Store: FAISS for semantic search
• —Framework: LangChain for document processing

Deployment

This app is designed to run on Hugging Face Spaces with Streamlit. It automatically handles:

• —Model caching for faster subsequent loads
• —Memory-efficient document processing
• —Error handling and graceful fallbacks

Limitations on HF Spaces

• —File uploads are limited to reasonable sizes
• —Models are cached but initial load may take time
• —Persistent storage is limited

Local Development

To run locally:

pip install -r requirements.txt
streamlit run app.py