Spaces:
				
			
			
	
			
			
		Sleeping
		
	
	
	
			
			
	
	
	
	
		
		
		Sleeping
		
	| title: cutomer support | |
| # π§ Customer Support Chatbot (FastAPI + Hugging Face + Qdrant) | |
| This is an intelligent, conversational customer support chatbot built using: | |
| - π **GLiNER** for named entity extraction | |
| - π§ **SentenceTransformers** + **Qdrant** for semantic search and retrieval | |
| - π€ **Groq LLM** for answer validation and response generation | |
| - β‘ **FastAPI** for the web backend | |
| - πΌοΈ **Jinja2 templates** for a simple frontend chat UI | |
| It runs both locally and on **Hugging Face Spaces via Docker**. | |
| --- | |
| ## π Features | |
| - Extracts key entities from support requests | |
| - Semantic similarity search with Qdrant + SentenceTransformers | |
| - LLM validation of results via Groq API | |
| - Fully functional FastAPI-based web interface | |
| - Docker-ready for Hugging Face Spaces deployment | |
| --- | |
| ## π¦ Installation | |
| ### Local Setup | |
| ```bash | |
| git clone https://github.com/your-username/support-chatbot | |
| cd support-chatbot | |
| python -m venv venv | |
| source venv/bin/activate | |
| pip install -r requirements.txt | |
| uvicorn main:app --reload |