BSc Computer Science (Cyber Security) student at APU/DMU. SOC analyst, penetration tester, digital forensics practitioner — and builder of tools across security, AI simulation, and infrastructure.
CLI TCP SYN scanner across CIDR ranges with banner grabbing, CPE mapping, and NVD CVE API lookups. Outputs standalone HTML + JSON reports with CVSS scores. Falls back to TCP connect when raw packets unavailable.
Privacy-first credential audit tool. zxcvbn strength scoring, entropy calculation, and HIBP k-anonymity breach checks — only the first 5 hex chars of the SHA-1 hash are sent to HIBP. Bulk audit endpoint, generator, React UI with animated strength meter.
Real-time quantitative market scanner and risk terminal for MetaTrader 5. Normalized weighted probability model (0–100 score) using EMA trend, RSI momentum, Donchian structure, ATR volatility, and spread liquidity. FTMO/5ers prop-firm compliance built in.
Room-based collaborative editor with OT conflict resolution, Monaco Editor, Redis-backed document state, debounced autosave, per-user cursors with colors, live chat, and language switching (JS/Python/Go). Jest tests for OT logic. Docker Compose setup.
Full-stack CI/CD monitoring dashboard polling GitHub Actions API every 60s, storing run history in SQLite, and exposing key-authenticated metrics endpoints. Slack Incoming Webhook alerts on pipeline failure. Automated CI via GitHub Actions.
Multi-agent simulation where agents (random, greedy, hazard-avoider, energy-saver) share a grid. Adaptive controller changes resources, hazards, and safe zones based on collective performance metrics. JSONL logs, metrics plots, MIT licensed.
Signal Collector game wrapped as a Gymnasium-compatible RL environment. YAML-driven difficulty schedule (easy → expert), optional moving hazards, dict observations, terminal renderer, and random-policy evaluation examples. Clean game/env separation.
Simulation environment designed around human-adaptive logic, extending the adaptive RL testbed series with interfaces oriented toward human-in-the-loop or human-behavioral modelling scenarios.
Runtime task engine that dynamically adjusts task parameters and difficulty at execution time based on performance feedback — part of the adaptive simulation infrastructure series.
Gymnasium-compatible gridworld environment with adaptive difficulty scaling. Complements the game wrapper and multi-agent playground to form a complete adaptive RL testbed suite.