AMR-Adapt: Evidence Synthesis & Bayesian Regimen Adaptation Tool
A solo-developed interactive web prototype for antimicrobial resistance (AMR) data analysis, designed to streamline evidence synthesis and support probabilistic treatment decision-making.
The tool automates literature ingestion and screening with real-time PRISMA flow visualization, computes sample-weighted pooled resistance estimates, and applies Bayesian modeling to rank AWaRe-classified empiric regimens. Built using Dash, Plotly, SQLite, Pandas, and SciPy, it processed 500+ record batches and reduced manual screening time from hours to minutes, demonstrating a complete pipeline from raw data to evidence-based treatment recommendations.
Explore Dashboard →