Inside the Network

Our guest in this episode is Snehal Antani, co-founder and CEO of Horizon3, one of the fastest-growing companies in offensive cyber that recently crossed the milestone of $100M ARR.

Snehal's journey of becoming a cybersecurity founder is anything but conventional. Before starting Horizon3, he held senior technology leadership roles at IBM, GE Capital, and Splunk, where he served as CTO. He then made an unusual career decision: leaving the corporate world to become the first CTO of the U.S. Joint Special Operations Command (JSOC). Working alongside some of the military's most elite teams fundamentally changed how he thought about leadership, technology, and solving problems under pressure.

In 2019, Snehal left JSOC to start Horizon3 with a team that combined special operations veterans and experienced enterprise software engineers. Their vision was to fundamentally change offensive cybersecurity by making penetration testing autonomous, continuous, and safe to run in production. Rather than relying on consultants or theoretical vulnerability scores, Horizon3's NodeZero platform demonstrates which weaknesses attackers could actually exploit.

On Inside the Network, Snehal shares the story of building Horizon3, raising capital without previous fundraising experience, and scaling through the dramatic market shifts of 2021 and 2022. He talks candidly about growing the company from 25 to 125 employees in just a few months, the cultural and operational problems that followed, and why startups must eventually transition from hiring “pirates” who thrive in chaos to building a disciplined “navy” capable of executing at scale.

We also dive into the future of AI-powered cybersecurity. Snehal explains why he believes the next era of cyber warfare will be AI versus AI, operating at machine speed, with humans involved by exception. 

Snehal offers a provocative perspective on the AI security startup landscape. We discuss why he believes many AI penetration testing startups are little more than wrappers around frontier models, why the most valuable cybersecurity training data lives behind enterprise firewalls, how deception could become one of the most effective defenses against AI-powered attackers, and why even the best cybersecurity technology can fail without strong distribution.