Back to Projects

HealthcareAI

LIVE

An AI-assisted healthcare platform featuring a RAG-based MediBot, NLP symptom analysis, and XGBoost disease prediction.

HealthcareAI

Overview

HealthcareAI bridges modern web architecture with machine learning integration. It provides a robust patient-doctor ecosystem, featuring a symptom-search module backed by NLP and an XGBoost prediction service, alongside a RAG-based MediBot to offer preliminary health insights before consultations.

Why I built this

To create an accessible entry point for patients to evaluate their symptoms and immediately route them to appropriate medical professionals within the same platform.

Engineering Challenges

  • Integrating a separate ML prediction microservice with the main Node.js backend
  • Structuring symptom search to be fast and forgiving of typos
  • State management across complex patient booking flows

What I learned

Decoupling the ML inference engine from the core API prevented long-running predictions from blocking standard web requests.

Stack

ReactReactNode.jsNode.jsExpressExpressMongoDBMongoDBRedisRedisXGBoostNLPRAGREST APIs