RAD Security

Here is a problem that sounds boring until it ruins your quarter: security teams get too many alerts. A modern cloud environment - containers spinning up and down, Kubernetes rescheduling workloads, identities assuming roles - generates an enormous volume of signals, most of which mean nothing. The industry's response, for about fifteen years, was to build better dashboards. Bigger screens, more colors, prettier ways to look at the pile of things you cannot possibly read.

RAD Security's bet is that the dashboard was never the answer. The company, founded in San Francisco in 2021, thinks the fix is to put an AI in front of the pile - one that reads each alert, decides whether it matters, and, within guardrails, does something about it. They call these workers RADBots. They have names and job titles. There is VulnBot, which handles posture. CloudBot does detection and response. GRCBot writes the compliance reports nobody wants to write. And Ask RADBot answers the follow-up questions, which is to say it is the one you actually talk to.

This is a slightly absurd framing, and RAD Security seems to know it, which is part of the appeal. But underneath the naming convention is a real technical argument. RAD does not start from a list of known-bad signatures - the digital fingerprints of attacks somebody has already seen and catalogued. It starts from behavior. Using eBPF, a Linux kernel technology that lets you observe what a workload actually does at runtime, RAD builds a fingerprint of normal - the processes a container spawns, the connections it makes, the way an identity typically behaves - and then watches for deviation. The pitch is that you catch the attack nobody has named yet, because you are not looking for the attack. You are looking for weird.

The company was not always called RAD. It started life as KSOC - Kubernetes Security Operations Center - a name that told you exactly what it did and exactly how narrow it was. In March 2024 it rebranded to RAD Security, and the rename was a strategy rather than a logo refresh. KSOC's Kubernetes posture product stuck around, but the company had expanded into identity threat detection, zero-day detection in the software supply chain, and behavioral response across the whole cloud footprint. When your scope outgrows your name, you either rename yourself for where the threat surface is going or you spend the next three years explaining that you do more than the name says. RAD chose the former.

The people behind it are a useful pairing. Brooke Motta, the CEO, spent her career on the revenue side of security companies - Rapid7, Bugcrowd, Wallarm, Sonatype - including a stint as chief revenue officer. Jimmy Mesta, the CTO, is a security engineer who has secured some of the largest containerized environments in the world and done the unglamorous work of building cloud security programs from scratch. Go-to-market plus deep technical is a common thing to claim and a rare thing to actually have. The market seemed to notice: in 2024 RAD was the only company to reach the finals of both the RSA Conference Innovation Sandbox and the Black Hat Startup Spotlight, winning the latter's audience award. Judges do not hand that out twice by accident.