Neomen User

I've built AI security, fraud detection, and computer vision systems designed for real-world deployment. My work combines machine learning with FastAPI, React, cloud platforms, and edge AI.

During my AI & Machine Learning internship at Inventeron Technologies and Business Solutions LLP, I built an end-to-end fraud detection system processing 50K+ simulated daily transactions. It achieved 96.8% ROC-AUC, 87ms latency, 31% better fraud capture, and an 87% reduction in alert investigation time through SHAP explainability.

I created NeuroSentinel, an AI security platform for LLM agents with dual-layer compromise detection, Graph Neural Network propagation tracking, FastAPI microservices, Redis persistence, Docker deployment, and a published PyPI SDK. The platform achieved 90.86% detection accuracy, 99.9% uptime, and 300ms response times.

I've also built an AI-powered Gmail client, an investment and sales forecasting platform, and a YOLOv8 wildlife monitoring system on ESP32-CAM. I'm currently pursuing an Integrated M.Tech in Computer Science Engineering and enjoy turning AI models into practical, production-ready applications.