Praneeth Penuganti

Led a comparative analysis of PaDiM and PatchCore, improving anomaly detection accuracy up to 99.6% AUROC on the MVTec AD dataset; collaborated with cross-functional teams to interpret model outputs and support decision-making on data quality in industrial contexts. Optimized the computational performance of PatchCore by reducing memory bank size 100× via coreset subsampling, cutting inference time while maintaining accuracy; contributed to anomaly detection efforts tied to data validation and system reliability assurance.

Optimized cataloging workflows using SQL and Excel for better internal analytics. Conducted brand validation audits and identified trends to improve data quality by 20%. Enhanced productivity and accuracy by 15% with improved quality checks.

Managed 100+ architect accounts and supported HNI clients throughout the customer lifecycle. Built dashboards to analyze store visits and assess the impact of multichannel communication.