GenAI Security: 99% Is Not Protection

Evgeny Kokuykin grew up in Tutayev, studied applied mathematics in Yaroslavl, and began as an intern in Diasoft's new office. Banking systems gave him an engineering base, while the growing office made him a senior contributor and informal lead. Leaving and returning helped him choose his next direction deliberately.

In international custom development, Evgeny moved from engineer to executive. A medical startup showed how personal contact and reliable delivery turn client distrust into access to a critical backend. At Akvelon he learned sales, marketing, team development, and the economics of a services business.

After the disruption of 2022, the team lost familiar clients and created Raft. They had to rebuild reputation, sales, and relationships in a different market. Machine-learning experience enabled a generative AI bet: first through experiments and visible expertise, then through use cases where implementation could produce real returns.

His security path grew from HIPAA reviews, a university course, and OWASP work on GenAI application risks. An attack-research lab became HiveTrace, a product for observing and protecting agentic systems. Protection is never absolute, while AI improves engineering work only when people can verify its output with informed judgment.