pratikrai0101

Engineered a real-time safety app detecting gender-based threats with 95% detection accuracy, reducing false positives by 20% during testing across 3 public datasets. Implemented Solidity smart contracts logging over 1,000+ test events, enhancing traceability and tamper-resistance and introduced a 'Security Switch' privacy mode. Constructed a cross-platform app supporting 100+ concurrent users and integrated geolocation for live threat alerts. Optimised system efficiency by reducing memory footprint by 30%, ensuring performance in low-resource environments.

Architected an ML-powered budget system processing 500+ simulated financial scenarios and achieving an R² score of 0.92 on validation data; secured a ₹50,000 prize at HackGenX. Utilised AutoGluon + Scikit-learn to increase model training speed by 35% and simulate IMF datasets spanning 10+ variables. Formulated dashboards and fund-tracking workflows handling real-time updates for 2 user roles, ensuring transparency.