AI: Transformatie Learning – TechSaksham (by Eduskills, Microsoft, and SAP)

As part of the AI: Transformatie Learning program under TechSaksham, I developed an AI-powered virtual health assistant to evaluate blood donation eligibility based on health parameters.

Project: AI Virtual Health Assistant for Blood Donation Eligibility
Duration: Jan 15, 2025 – Feb 15, 2025
Technologies: Python, Streamlit, Scikit-learn, Random Forest, Pandas, NumPy, Imbalanced-learn

  • Machine Learning Model: Built a Random Forest classifier using 1,000+ synthetic health records generated with custom rule-based labels.
  • Web Interface: Developed an intuitive Streamlit app for real-time eligibility checks based on inputs like age, weight, BP, and conditions.
  • Oversampling: Used RandomOverSampler to handle class imbalance and improve accuracy.
  • Chatbot Integration: Added a rule-based NLP chatbot to answer common queries like low hemoglobin tips.
  • UX & Explainability: Provided clear reasons for ineligibility and personalized health suggestions to enhance user trust.
TechSaksham Internship Certificate