Pardon Mukoyi did not arrive at offensive security through any of the usual doors. There was no graduate scheme, no four-letter consultancy on the CV, no Ivy-League pipeline. There was a curiosity that started early, a discipline that came hard, and a sequence of vulnerabilities — submitted to the right programs at the right time — that turned a name into a reputation.
By his mid-twenties, that reputation had earned him acknowledgements from Microsoft, the European Union, Red Hat, Bank of America, Nike, AT&T, Achmea, UN Women, and a long list of organizations whose names sit behind NDAs. His work has been credited in over thirty public security disclosures, with bounty earnings crossing the million-dollar mark across the world's most demanding programs.
"The interesting work isn't in the tools. It's in the gap between what the tool will find and what an attacker will actually do."
Before founding Cyber Matrix, he had already been working at the edge of AI security — long before the industry agreed it was a category. Independent research into model behaviour, adversarial inputs, and the security implications of agentic systems became the foundation for what Cyber Matrix now calls Offensive Neural — a category-defining capability the firm operates for a small portfolio of organizations.
The firm itself was founded in Sandton in 2018. It has stayed deliberately small, deliberately selective, and deliberately quiet. The work it has done for some of the world's most consequential organizations is, by design, work the public will never read about.
What is on the record is enough.