tgTrax: Telegram User Activity Tracker & Correlation Analyzer - Monitor online patterns, detect user relationships, and visualize temporal behavior

Privacy-first behavioral intelligence framework for multi-platform analysis and sociodynamic research.

A complete machine-learning system that predicts AI assistant user satisfaction using behavioral signals such as device, usage category, time features, session metrics, and model metadata. Includes full ML pipeline, SHAP explainability, evaluation suite, and an interactive Streamlit analytics dashboard.

Simple vispy-based frame-by-frame behavior annotation GUI. Also can display DeepLabCut poses within videos during annotation.

Behavioral Signals SDK - REST and Streaming APIs for Behavioral Processing and Deepfake Detection

A tool to check the reliability and to make some analysis of data from Live Mouse Tracker (LMT) experiments.

Malware sandbox for automated PE/ELF analysis with EDR integration and behavioral monitoring. Open-source alternative to CAPE sandbox.

CupaX is an open-source malware analysis platform designed for security researchers, students, and defensive security teams. It automatically executes suspicious files in an isolated Windows environment and captures detailed behavioral analysis.

Multi-agent simulation exploring how AI personality traits affect cooperation and information sharing between kingdoms during crisis scenarios using LLMs

Post-DeepLabCut Coordinate Analyzer for Arena Behavior Experiments

KeyScope is a cross-platform, self-installing keylogger with active window tracking, developed for cybersecurity internship research and lab-based red team simulations. Intended strictly for ethical and educational use.

Automated video analysis tool for mouse behavioral experiments (Open Field, Social Avoidance, Forced Swim Tests).

Behavior-based Linux malware profiler: ptrace syscall tracing → JSONL events → heuristic analysis with IOC extraction & MITRE ATT&CK mapping. No signatures, no hashes

Turn-by-turn detection of manipulation accumulation in multi-turn AI conversations. Built for APART Research AI Manipulation Hackathon 2026.

Early warning SaaS churn prevention system using behavioral drift detection.

Open-source computer vision framework for 3D motion tracking and deep behavior analysis.

ML model training pipeline for the main Spyware Detector project. Dynamically serves updated models via GitHub Releases. Includes Dockerized training, automated CI/CD, and versioned artifacts (~97% accuracy). Educational project - not for production use.

Obfuscation-resistant YARA rule generation with deep archive scanning and behavioral analysis.

Real-Time Suspicious Activity Detection through Behavioral Modeling leverages machine learning to identify anomalies by modeling normal user/system behavior. It enables real-time threat detection, fraud prevention, and security monitoring by detecting deviations from expected patterns.