A real-time analytics tool for law enforcement to detect suspicious patterns in massive IPDR logs using Kafka and Flink. Developed for a cybersecurity hackathon to aid in digital forensic investigations.
Designed a secure, AI-based system for real-time fraud detection and behavior analysis. Utilized knowledge distillation, time-series RNNs, and explainability tools (LIME) to ensure efficient and transparent decision-making.
Developed a Rule-based Network Intrusion Detection (NID) system incorporating real-time packet inspection, slashing network breaches by 15% and flagging 200+ suspicious packetsdaily.