Humanix – Human Threat Detection & Response for Social Engineering Attacks. Welcome to boldstart.
At boldstart, we partner from Inception with bold technical founders reimagining the enterprise. As AI accelerates social engineering attacks, one of the most urgent gaps in cybersecurity has become clear: we’ve spent decades protecting systems while ignoring humans where 76% of breaches actually start.
That’s why we’re excited to announce our Inception investment in Humanix, led by Keith Stewart. We led the initial round alongside Evolution Equity, and we doubled down in the Series A led by Acrew Capital as Humanix launches the industry’s first Human Threat Detection and Response platform.
Stop blaming humans. Start defending them.
76% of breaches succeed by targeting people, not systems (Verizon DFIR 2024). The industry’s response? An avalanche of security awareness training. We’re blaming victims while attackers use conversational AI to manipulate employees in real-time across voice, video, email, and chat.
Traditional security stops at the network edge. When an attacker is socially engineering someone into wiring money, defeating MFA through the help desk, or approving a suspicious request, existing tools go completely blind. We’ve spent billions protecting systems. It’s time to protect humans.
There’s signal in the conversation. AI can detect it.
Keith’s insight came from listening to the Scattered Spider attack recording; a single phone call to a help desk that led to a $100M loss. The manipulation, pressure, and policy evasion were all there in the conversation. We can’t expect humans to catch these attacks 100% of the time, but we can train AI to detect them. Social engineering is an attack class like any other, and instead of exploiting code, it exploits natural language.
Humanix built the defense.
Humanix uses conversational AI trained specifically on human psychology and natural language attacks. The platform detects manipulation, deception, and impersonation as they happen across voice, video, chat, and email. If attackers weaponize AI to supercharge social engineering, we fight back with AI to protect our people.
The system continuously learns from each environment, adapting to detect subtle variations in tone, urgency, and manipulation tactics that humans miss and traditional tools can’t see. Humanix gives security teams visibility and control over their human attack surface for the first time.
Simple to deploy, impossible to ignore.
Humanix plugs directly into existing systems like Microsoft 365, Slack, ServiceNow, and Zoom with just an API key. No rip-and-replace. No months-long integrations. Just immediate visibility into social engineering attempts targeting your people.
We first met Keith in 2023 as AI was fundamentally reshaping the threat landscape. With 20+ years building security products, including as SVP of Product and Strategy at vArmour (where he later served as interim CEO through their exit), Keith had the technical chops and product instincts to see what the industry missed: while everyone was adding AI features to existing tools, nobody was defending the humans that attackers actually target.
Keith and Humanix are building a new defense layer for an attack surface that’s been wide open for decades. The product is live in production environments today, and we’re excited to partner with him from inception as he builds what we believe will become a foundational category in cybersecurity.
Welcome to the boldstart family!