Nadheem Muhammed

Architected a mobile clinical decision-support tool that used OCR (Tesseract) and vector search in a database (ChromaDB) to deliver context-grounded drug information to clinicians. Performed clinical validation to identify reasoning failures and eliminate AI hallucinations by verifying outputs against real FDA package insert information. Presented digital health architecture and clinical workflow models to institutional stakeholders via KSUM.

Extracted and governed longitudinal ICU datasets from electronic hospital archives, verifying data integrity across clinical markers. Built supervised ML classification models in Python to predict pathogen transmission trends and evaluated results using AUC-ROC curve analysis. Focused on data quality checks and traceable, governance-aligned dataset handling for research outcomes.

Conducted clinical rotations across ICU, Oncology, and Medicine, evaluating bedside documentation, patient progress notes, and lab panels. Audited pharmacotherapy regimens to support antimicrobial stewardship protocols and verify compliance with clinical guidelines. Managed adverse drug reaction surveillance and regulatory safety reporting workflows aligned with WHO Vigiflow standards, ensuring structured documentation for pharmacovigilance and data safety.