The "move fast and break things" founder archetype just got empirically debunked by a 46-year-old who spent two decades learning before launching.

The Summary

The Signal

Antani's company builds AI agents that attack your infrastructure before real hackers do. Horizon3's autonomous penetration testing platform runs continuous simulated breaches, finding zero-day vulnerabilities and misconfigurations that human security teams miss. This is the agent economy applied to cybersecurity: software that thinks like an attacker, operates 24/7, and scales across thousands of attack vectors simultaneously.

The timing matters. As companies deploy more AI agents to handle customer service, data analysis, and workflow automation, the attack surface explodes. Every agent is a potential entry point. Every API connection is a door that might be unlocked. Traditional security audits, done quarterly by humans, can't keep pace with systems that evolve daily.

"I wanted to make sure I built the right experiences before I built a company."

Antani's preparation checklist reads like a deliberate curriculum:

  • Product development at IBM
  • Organizational management at GE Capital
  • Hypergrowth scaling at Splunk
  • Leadership under pressure at Joint Special Operations Command

That last one is where he met cofounder Anthony Pillitiere, an Air Force veteran. Military cyberwarfare experience isn't decorative for a company simulating attacks. It's domain expertise in adversarial thinking, red team operations, and systems under active threat.

The double unicorn valuation puts Horizon3 in rare company. Only a handful of cybersecurity startups have crossed $2 billion while still private. The $250 million raise suggests investors see AI-powered offensive security as critical infrastructure for the agent economy. You can't deploy autonomous systems at scale without autonomous systems testing their security.

The Implication

The "young genius founder" narrative persists because it's better content than "middle-aged expert applies two decades of knowledge." But if you're building in AI security, infrastructure, or anything touching regulated industries, Antani's path is the template. The hard problems require domain depth, not just coding speed.

Watch for more AI security startups built by people who spent years in offensive cyber roles, either military or big tech red teams. The agents coming online in 2027 need guardians who think like attackers. That's not a skill you pick up in a hackathon.

Sources

Business Insider Tech