Flagship work / Research / Get in touch
I'm Hadi, a cybersecurity engineering student working where security, machine learning, and systems engineering meet. I build the full path: capture attacker behavior, analyze the evidence, and give an investigator something useful to work with.
Open to cybersecurity internships and entry-level roles. My projects are open source, designed for research and authorized testing, and built to run on infrastructure I control.
HoneySentinel AI captures attacker interactions across SSH, Telnet, FTP, HTTP, and HTTPS and turns them into encrypted evidence. Its investigation workspace brings together session transcripts, ATT&CK mappings, indicators, and exports for downstream tools.
The self-hosted deployment blocks outbound traffic from the honeypot engine and separates ingestion from the API. Wolfram enriches captured transcripts through local inference.
Explore the system → · Architecture & deployment · Evaluation & measurements
Wolfram is the final evolution of my Chimera project. Built on MiMo-V2.6-Distill-Qwen-9B, it is fine-tuned for security analysis: what an attacker intended, which objectives and ATT&CK techniques the commands support, and which indicators appear in the transcript.
The project covers fine-tuning, evaluation, GGUF deployment, and local integration with HoneySentinel. The application validates technique IDs and checks extracted indicators against captured transcripts.
Explore Wolfram → · See the integration
Systems & security — Python · Go · C/C++ · Linux · Docker · Protocol emulation · MITRE ATT&CK
Models & agents — PyTorch · Transformers · QLoRA · llama.cpp · Hybrid RAG · MCP
Applications & data — TypeScript · React · FastAPI · PostgreSQL · Redis · SQLite
