MemoryKeeper is an AI-powered assistant designed to help individuals with Alzheimer's disease recall important memories, facts, and daily information.
The bot uses natural language processing (NLP) techniques to retrieve personalized information and respond to user queries in a supportive, friendly manner.
This project aims to offer cognitive support and memory stimulation using a Retrieval-Augmented Generation (RAG) architecture.
- Conversational memory support for Alzheimer's patients
- Information retrieval from a custom knowledge base
- Natural language understanding and friendly responses
- Easily extensible for personalized memory datasets
- Python 3.11
- Hugging Face Transformers
- FAISS (for vector search)
- Scikit-learn
- Gradio (for demo app)
- Personalized memory updates based on user feedback
- Voice assistant integration
- Mobile-friendly version
- Contextual memory chaining