Behavioural Analysis Tool

Unified object detection and gaze intersection tracking for cognitive science research

Map the chaos of online conversations into structured, understandable narratives.

Context-aware phishing detection system that analyzes URL structure, behavioural risk signals, and multilingual social-engineering patterns to identify phishing attempts.

Python based cybersecurity tool that detects suspicious processes and potential keylogger activity using real time monitoring and rule based analysis.Detect suspicious processes, potential keyloggers, and hidden threats running on your system in real time..

Behavioral security layer for NVIDIA OpenShell — patent-pending DFS-based agent behavioral analysis

Limbless Animal traCkEr - a markerless, contour-based pose estimator for limbless animals (MATLAB). Garg et al. 2022, Front. Behav. Neurosci.

Multi-Camera Synchronization and Instance Tracking for Avian Behavior Analysis in a Floating Aviary

Computer vision pipeline for automated behavioural analysis from video — body pose estimation and facial expression classification using MediaPipe.

Pose and behavioural metric extraction from video via MediaPipe — head orientation, wrist velocity, interpersonal distance.

AI-driven burnout assessment and recovery platform operationalising the NEU Framework through behavioural science, analytics, and personalised intervention systems.

Productivity Stability as a Behavioural Marker of Self-Regulation from Renzo Pomodoro

An interactive web application for exploring the psychometric battery results of Qwen 2.5 0.5B Instruct.

Behavioral analytics engine that processes raw user logs into "Year-in-Review" visual narratives (Spotify Wrapped style) using Matplotlib and Seaborn.

An Isolation Forest model that detects simulated hackers in server logs.

This project presents a data-driven analysis for Nature's Basket, a modern grocery delivery service operating in Mumbai, offering fresh fruits, vegetables, snacks, and beverages. The company aims to enhance customer satisfaction by addressing challenges related to order cancellations and by better understanding customer behavior.

Python-based Sysmon XML parser and security analysis pipeline with rule-based detection, behavioral correlation, MITRE ATT&CK mapping, risk scoring, reports, and optional OSINT enrichment.

Provide a Stated Preference (SP) Survey analysis focusing separately on experimental design, questionnaire creation and SP econometrical analysis

Production-style fraud detection pipeline for financial transactions using behavioural analytics, anomaly detection, and machine learning with out-of-time validation and explainable risk scoring.