When JetStream Security closed its seed round earlier this month, the company wasn't looking to solve every problem in enterprise AI. It was, in essence, aiming at one: the fact that most companies deploying artificial intelligence have little idea what their systems are actually doing.
The Santa Clara startup announced the $34 million round on March 3, led by Redpoint Ventures with backing from the CrowdStrike Falcon Fund. What caught attention wasn't just the size—substantial for a seed deal—but the roster of angels who joined. CrowdStrike CEO George Kurtz. Wiz CEO Assaf Rappaport. Okta Vice Chairman Frederic Kerrest. According to the company, the round wrapped "in a matter of weeks" and drew interest well beyond what it could accommodate.
That kind of velocity suggests something: either remarkable founder pedigree, or acute market timing. JetStream appears to have both.
Veterans of the Endpoint Wars
CEO Raj Rajamani spent the better part of a decade building products at two companies—SentinelOne and CrowdStrike—that helped define how enterprises think about endpoint security. He was Chief Product Officer at CrowdStrike from 2023 through 2025, and held the same title at SentinelOne from 2017 to 2022.
His co-founders trace similar arcs. COO Jared Phipps was most recently CRO at Dazz after years in senior roles at SentinelOne. CTO Jatheen "AJ" Anand logged time at Lightning AI and Khoros in VP engineering roles. Chief Architect Venu Vissamsetty was there at the start of Attivo Networks, later moving into a VP Engineering seat at SentinelOne.
Which is to say: this is a team that knows how to sell security to large, risk-averse organizations. And they've watched those same organizations begin deploying AI agents—chatbots, autonomous workflow tools, code generators—without a clear map of what those agents can access, where they're pulling data from, or who granted them permission in the first place.
"Leaders are being asked to bet their businesses and careers on systems they can't fully see, explain, or control," Rajamani said in the company's announcement. It's not an unfamiliar pitch in enterprise software. But it's resonating differently now.
Blueprints, Not Dashboards
At the core of JetStream's platform is something the company calls AI Blueprints. Think of them less as static documentation and more as live, constantly updating maps of how AI systems operate inside an organization. Which models are calling which APIs. Which agents have access to sensitive customer data. Which workflows were approved by IT, and which sprouted up in a sales team somewhere without anyone's knowledge.
The system attempts to stitch together visibility across SaaS applications, endpoints, cloud environments, and APIs—a wide surface area that reflects just how fragmented AI deployments have become. JetStream pairs that visibility with what it describes as "design governance," essentially ways to structure agentic systems before they go live, and "runtime governance," which flags when a live agent starts behaving differently than its approved blueprint suggested it would.
There's also an identity layer (the company calls it "agentic identity management," with attribute-based access controls baked in) and an AI FinOps module to track which models and agents are burning through API credits the fastest.
The underlying thesis is straightforward, if a bit grim: enterprises are creating "shadow AI" sprawl at a pace that makes the old problem of shadow IT look quaint. And unlike rogue SaaS subscriptions, rogue AI agents can do more than waste budget—they can leak proprietary data, make customer-facing decisions no one authorized, or simply break in ways that are hard to diagnose after the fact.
Whether the market will treat AI governance as a distinct category—separate from security, compliance, or observability—remains an open question. But investors seem willing to find out.
A Pattern, Not an Outlier
JetStream isn't placing this bet in isolation. Protect AI, another startup in the space, raised $60 million in August 2024. A company called 7AI reportedly closed a $130 million Series A late last year, though details on that round have been sparse. And a stream of research—from EY in June, from Gartner around the same time, from Operator Collective just last week—keeps painting the same picture: adoption is running ahead of governance. Security concerns rank high among barriers to broader AI rollouts, particularly at organizations mature enough to worry about them.
An Operator Collective survey released by Axios on March 4 found that while experimentation with AI is widespread, few companies have moved past pilot stages or established clear ways to measure ROI. In other words, they're flying blind. Which is, more or less, the market condition JetStream is designed to exploit.
Redpoint's Erica Brescia pointed to the team's "track record" and the moment—agentic AI "reshaping operations," in her phrasing—as reasons the firm moved quickly. It's the kind of language VCs tend to use when a category is still forming but the founders have done it before.
What's Next, and What's Unclear
JetStream says the fresh capital will go toward hiring across engineering, product, and go-to-market. The company has already assembled an advisory council stocked with executives from Descope, CrowdStrike, and alumni of Attivo Networks and Zscaler—a setup that signals channel partnerships and enterprise sales motion ahead.
The company claims it's already working with Fortune 500 customers, though it declined to name them. That's typical for early-stage enterprise startups, particularly in security, where customer logos can be slow to materialize publicly.
Still, there are questions. Governance platforms live or die on integration depth, and JetStream will need to prove it can plug into the sprawling, messy stack of tools enterprises actually use—not just the tidy architecture diagrams in pitch decks. The market is also crowded with observability, security, and compliance vendors who may argue they already do some version of what JetStream is describing, even if they don't call it "AI governance."
And then there's the broader uncertainty: whether enterprises will truly treat agentic AI governance as a must-have category, or whether it gets absorbed into existing procurement cycles as a nice-to-have feature from an incumbent vendor.
For now, though, JetStream has time, capital, and the kind of founder credibility that opens doors. In a market where "shadow AI" is starting to sound less like a buzzword and more like a boardroom liability, that might be enough.
