About The Project
MedPrompt+ is an advanced retrieval-augmented generation (RAG) platform tailored for clinical decision support and personalized health advice. By securely parsing complex medical history documents and clinical notes using advanced OCR pipelines, it creates a searchable vector database. Patients and doctors can converse with an empathetic LLM backed by clinical evidence, predict health risks using trained machine learning models, and generate customized wellness schedules.
Key Features & Architecture
Hybrid RAG Architecture
Combines keyword search with vector embeddings for zero-hallucination medical chat
Advanced OCR Document Parser
Extracts tabular and unstructured data from patient health records
ML-Based Risk Predictor
Leverages scikit-learn classifiers to screen for cardiovascular and diabetic risks
Contextual Wellness Planner
Automatically compiles dietary, exercise, and check-up schedules
Secure Patient Portal
Clean dashboard showcasing diagnostic history and health trend analytics