BM
Brooke Motta
Co-founder and CEO, RAD Security
Brooke leads the company and customer strategy behind RAD Security. AgentKeeper carries the same enablement lens: security should help teams adopt new technology with evidence and control.
About AgentKeeper
Built by the runtime security team behind RAD Security.
AgentKeeper exists because AI agents stopped being chat windows and started taking action on developer workstations. RAD Security has spent years turning runtime behavior into usable security signal. We are bringing that discipline to AI adoption.
Founding line
RAD Security -> AgentKeeper
origincloud and workload runtime behavior
patternobserve, evaluate, enforce, audit
new surfaceAI agents on developer workstations
productAgentKeeper
Runtime security roots
RAD Security
Built from years of cloud-native and behavior-driven security work, not a prompt-scanning wrapper.
Founders
Brooke + Jimmy
A company-builder CEO and a technical founder known for cloud-native security depth.
Series A
$14M
Led by Cheyenne Ventures, with security and platform investors around the table.
Recognition
RSA + Black Hat
RAD Security was selected for both major 2024 startup competitions and won Black Hat audience choice.
Why we are here
AI agents need the same discipline as production systems.
The first wave of AI adoption asked security teams to approve tools they could not see. AgentKeeper gives those teams a way to enable the tools and keep policy close to the action.
2022
RAD Security raises its seed round
The company starts with a clear runtime thesis: security should understand what workloads actually do, then turn that behavior into action.
2024
Behavioral defense gets market signal
RAD earns recognition at RSA Innovation Sandbox and Black Hat Startup Spotlight as security teams look for cleaner signal in cloud and AI environments.
2025
RAD expands into AI-driven defense
A $14M Series A backs the next phase: real-time defense for cloud and AI systems, with visibility, automation, and investigation built in.
Now
AgentKeeper brings that model to developer agents
AI coding agents moved the risk from infrastructure alone to tool calls on workstations. AgentKeeper puts policy in that path.
monitor -> enforce -> audit
Founding team
Security operators, company builders, and runtime engineers.
JM
Jimmy Mesta
Founder and CTO, RAD Security
Jimmy is a veteran security engineering leader focused on cloud-native security. His background includes large-scale cloud security programs, technical security training, research, and securing major containerized environments.
Backed by security people
Investors who understand infrastructure, security, and adoption.
RAD Security has been backed by institutional investors and operator angels across cybersecurity, cloud, and developer infrastructure. AgentKeeper is built inside that same company.
Venture backers
Cheyenne Ventures
Forgepoint Capital
Lytical Ventures
Akamai
.406 Ventures
Vertex Ventures US
Gula Tech Adventures
Operator angels and mentors
Andrew Peterson, Founder, Signal Sciences
Carol Meyers, Former CMO, Rapid7
Jim Manico, Founder, Manicode
Grant Miller, Founder, Replicated
Why we are qualified
We are not learning security from the agent era. We are bringing security into it.
We know runtime behavior
RAD's work is grounded in behavioral runtime insight. AgentKeeper applies that same idea to the moment an AI agent reads, writes, runs, fetches, or calls a tool.
We know enterprise adoption
The team has built for security buyers who need to say yes without losing the audit trail. AgentKeeper starts in observation and moves to enforcement as teams are ready.
We know security signal
The goal is not more alerts. It is a policy decision with context: user, workstation, agent, tool, arguments, verdict, and evidence.
What we believe
The right control makes adoption faster.
AI adoption should not depend on blind trust.
Policy should live where the agent acts.
Security teams need evidence, not theater.
Developers should keep useful tools without hiding from governance.
Built by RAD Security
Enable AI agents with runtime evidence in the path.
Start with one workstation, prove what the agents are doing, then scale the rollout with policy your security team can trust.