Michał Suchocki
Warsaw · Offensive security & AI

Michał
Suchocki.

dr inż. CEO & co-founder, CyCommSec creator of FuseAI ISO 42001 lead auditor
Thesis

I like being there for version one.

I started at Samsung, taking the first Galaxy phones through field testing before anyone knew whether the line would matter. Then Huawei, as a product manager on one of the first commercial LTE networks in Poland — at a point when LTE was still mostly a promise on a slide. At Cybercom I built the telecom, automotive and IoT practices, shipped platforms on three continents, and eventually ran the region out of Dubai. I co-founded Cyberteq and grew it across the Middle East and Africa. In between there was technology strategy at Accenture, consulting at Detecon, and several years advising Philip Morris International as a security and generative-AI architect.

I came into all of it as an engineer. I trained in electronics and telecommunications at the Military University of Technology in Warsaw and stayed long enough to finish a doctorate — on the human auditory system, classifying brainstem evoked potentials with wavelet decomposition and support vector machines. The point was to let a machine read a neurological signal a clinician would normally interpret by eye, and read it more consistently. That software is still used in hearing diagnostics, and it became Neurality. So before I ever automated an attack I had spent years teaching algorithms to find the signal in the noise. The problems turned out to be more alike than they look.

Today I run CyCommSec in Warsaw, and I made FuseAI — a multi-agent system that uses large language models to automate penetration testing and the whole cycle of vulnerability management. I think offensive security is one of the places where AI will change the rules fastest, and I'd like to be one of the people who decides how. But cybersecurity is where I do the work right now, not what the work is. Building is what I actually do — and I expect the next thing I build will look nothing like the last one.

I still teach: penetration testing and OT security at AGH in Kraków, cybersecurity and the theory of computation and simulation at WAT, AI in cybersecurity at SGH. I write for Forbes Poland and speak when I can be useful. I work in Polish and English, from Warsaw.

Full biography →
Research
97%
Classification accuracy, clinically validated — 95% sensitivity, 98% specificity.
2018
Doctorate, Military University of Technology (WAT), Warsaw.
3
Universities where I lecture: WAT, AGH Kraków, SGH Warsaw.

My academic work is in biomedical signal processing — using machine learning to interpret neurological signals. The software it produced is still used in hearing diagnostics.

Classification of auditory brainstem response using wavelet decomposition and SVM network
Dobrowolski, Suchocki, Tomczykiewicz, Majda-Zdancewicz · Biocybernetics and Biomedical Engineering 36(2), 2016
Training

I teach this for a living, not as a sideline.

Three university lectureships and seventeen years of shipping the systems under discussion. The courses below are the same material, compressed for people who have to make decisions on Monday.

Selected talks
FuseAI: a multi-agent system for penetration-testing automation The Hack Summit · 2025
Cybersecurity: not only defence, but also attack CYBERSEC CEE · 2024
Cyber offense in risk management AGH Cyber Kampus · 2023