With backgrounds in top executive roles at major cybersecurity vendors including CrowdStrike and SentinelOne, the founders of JetStream tell CRN they are not looking for a quick exit for the startup.
Amid the surging demand for businesses to deploy AI agents for boosting workplace productivity, AI security startup JetStream Security is looking to build up an enduring platform with the help of channel partners, JetStream founders told CRN.
Founded in 2025—and having emerged from stealth in March—the startup has opted to work with solution and service provider partners from the get-go as the best way to scale up in this intense, AI-accelerated environment, the executives said.
[Related: Why Rogue AI Agents Are A Wake-Up Call For Security Teams: Experts]
With backgrounds of serving in top executive roles at major cybersecurity vendors including CrowdStrike and SentinelOne, the founders of JetStream said they are not looking for a quick exit for the startup. Instead, the goal is to develop an AI security platform that can expand over time alongside customer deployments of AI—while also building a company that channel partners can build a business around for years to come, the executives said.
In other words, while the JetStream platform is starting out by providing security teams with a way to see what AI agents are doing—while also restricting their access and helping to keep AI spending in check—JetStream’s ambition goes well beyond addressing individual AI security problems.
“We wanted to build something foundational, something lasting,” JetStream co-founder and COO Jared Phipps (pictured left) told CRN.
That approach is resonating with solution providers such as GuidePoint Security, No. 32 on CRN’s 2026 Solution Provider 500.
Herndon, Va.-based GuidePoint has already closed multiple deals with JetStream and sees a strong opportunity for the startup to become a significant cybersecurity player going forward, according to Mark Thornberry, chief business officer at GuidePoint.
“I think that the next great cyber company is going to be born now,” Thornberry said. “And I think that [JetStream] has as good a shot to capture that as anyone that I’ve come across this year.”
All in all, “it’s been great to have a front row seat to it, and I think we’ll continue to be a major part of that [growth],” he said.
During its exit from stealth in March, JetStream announced $34 million in seed funding led by Redpoint Ventures, with participation from the CrowdStrike Falcon Fund. Individual investors in the round included CrowdStrike co-founder and CEO George Kurtz, Wiz co-founder and CEO Assaf Rappaport and Okta co-founder Frederic Kerrest.
JetStream’s top leadership includes co-founder and CEO Raj Rajamani (pictured second from right), who previously served as chief product officer at CrowdStrike as well as at SentinelOne, along with Phipps, whose previous roles include leading sales and solution engineering at SentinelOne.
Other founders include CTO Jatheen “AJ” Anand (pictured second from left) and Chief Architect Venu Vissamsetty (pictured right).
AI Security, Cost Controls
The startup sees a massive opportunity to bring together capabilities that many businesses have been attempting to assemble themselves from various tools and open-source components, Rajamani told CRN.
JetStream’s platform ultimately combines AI security with governance and financial controls—and puts a key emphasis on helping organizations understand how their AI systems are operating, he said.
Crucially, JetStream also enables intervention within AI systems when necessary, according to Rajamani.
A central component is what JetStream calls its AI blueprint, which provides a view of an AI system—including across services, agent permissions and guardrails—that is updated dynamically, according to the company.
That visibility can go a long way toward supporting compliance controls, while also helping to detect when a system has begun to behave outside of expectations, Rajamani said.
For instance, JetStream can flag or block unauthorized activity, or require human approval before allowing the activity to continue, he said. The overall goal is to provide controls that go beyond basic instructions for governing how a model or agent should behave, according to the startup.
“Guardrails have their place. They have some value. But they’re very brittle, and they’re by no means [a be-all], end-all,” Phipps said.
The bottom line is that “we built a control plane, and we can say ‘yes’ and ‘no’ to agents accessing resources on a just-in-time basis,” he said.
Notably, that same ability to control access to resources also creates a huge opportunity to help businesses manage their AI spending, Phipps said.
JetStream’s financial operations capabilities help customers identify where funds and AI tokens are being consumed, as well as enforce budgetary limits, Rajamani said. Spending can ultimately be optimized over time with the system, he said.
Phipps said the decision to build those capabilities came directly from customers and partners.
“It didn’t come from us. It came from partners [and] customers saying, ‘Can you control that too?’” he said. “Nobody’s running JetStream only for that. But everybody that’s running JetStream is using that.”
Partner Perspective
For GuidePoint Security, the cost-control capabilities have clearly stood out as noteworthy in feedback from the solution provider’s engineers and technical leaders, Thornberry said. Without a doubt, helping CISOs and CIOs with understanding and managing AI costs is an increasingly important part of the conversation right now, he said.
The company’s approach also aligns with what GuidePoint is building through its own AI services, Thornberry said.
Currently, GuidePoint is working closely with JetStream in most regions of the U.S., and “we’re very bullish on where they’re headed with our partnership,” he said.
JetStream’s rapid channel traction at such an early stage of the startup’s growth has certainly stood out, as compared to prior companies that the company’s founders have been involved with, according to Rajamani.
“This type of traction, with such an early product and company, is something that I have not experienced,” he said. “Usually, it takes at least one or two years of priming the pump to get the channel excited [to this degree].”
Already at this stage, however, “we are able to go close very large deals with Fortune 500 accounts” with the help of partners, Rajamani said.
The speed of the early deals reflects massive customer needs that many partners had already been working to address, but without having found an ideal option previously, Phipps said.
Without question, businesses are increasingly turning to partners that already know their environments to address AI security challenges, he said.
As Phipps put it, “the trusted partners who’ve been architecting [and] building with these customers for decades—that’s who people are turning to for answers.”
