CLI red-teaming framework for LLM applications. Runs structured adversarial campaigns against models, RAG systems and agents using a three-layer attack taxonomy, then produces scored reports with verified metrics for reliable, reproducible LLM security audits.
OSINT Agentic Operations — multi-agent OSINT platform that automates open-source investigations and IOC (indicator of compromise) extraction, orchestrating specialised agents across each phase of the intelligence cycle.
Network Analysis Module — my final degree project (2020): a multithreaded C packet sniffer (pcap with BPF filters) that samples 15 traffic features, sends them over UDP and classifies them in real time with a Random Forest.
Automatic meal planner: generates balanced weekly menus and shopping lists from personal constraints and preferences.