For years, enterprise security teams have wrestled with a peculiar irony: the tools designed to prevent data breaches often create so much noise that real threats slip through unnoticed. Jazz, an Israeli startup that emerged from stealth mode in mid-March, believes it has an answer—one that's already attracted $61 million in funding and a handful of high-profile customers willing to bet on a different approach.
The company's proposition is deceptively simple. Instead of the rules-based systems that flood analysts with thousands of daily alerts—most of them false positives—Jazz deploys AI agents to investigate potential data loss incidents before they ever reach a human. The result, according to the company, is roughly 10 pre-vetted alerts per day. If that sounds too good to be true, well, that's the question Jazz now has $61 million to answer.
Founded by four veterans of Unit 81, Israel's intelligence technology unit, Jazz announced on March 10 that it had raised an $18 million seed round followed by a $43 million Series A. Both rounds were co-led by Glilot Capital Partners and Team8, with participation from Ten Eleven Ventures, Merlin Ventures, Encoded Ventures, MassMutual Ventures, and a handful of unnamed entrepreneurs from the cybersecurity and AI worlds. The company declined to disclose valuation figures or specify when each round actually closed.
From Military Intelligence to Enterprise Security
CEO Ido Livneh leads Jazz alongside three co-founders: Jake Tuertskey as Chief AI Officer, Noam Issachar handling business operations, and Yonatan Zohar serving as CTO. The team also includes alumni from Axonius and Laminar—companies that have tackled adjacent problems in cloud security and data governance. That pedigree matters in enterprise security, where trust is everything and a misstep can mean the end of a young company.
Today Jazz employs 45 people split between Israel and the United States, though recent job postings in Tel Aviv for backend, agent, and AI engineering roles suggest that headcount is climbing.
Rethinking an Old Problem
Data loss prevention isn't new. Traditional DLP systems work by defining rules—regex patterns that scan for credit card numbers, social security digits, proprietary file types—and flagging anything suspicious. The problem is that "suspicious" has become an almost meaningless category. Employees share documents constantly. Cloud collaboration tools proliferate. And with the rise of generative AI platforms, the potential pathways for sensitive data to leak have multiplied.
The result? Security teams drowning in alerts, most of which turn out to be benign.
Jazz's answer is what it calls "intent and context." Rather than simply pattern-matching, the company's system attempts to understand why data is moving and whether that movement aligns with business logic. An employee emailing a spreadsheet to a personal account on a Friday afternoon might be innocent (working from home over the weekend) or malicious (preparing to jump to a competitor). Context matters.
The technology stack includes an AI agent the company has dubbed "Melody"—the Agentic Investigator—that conducts autonomous investigations into potential incidents. A forensic endpoint agent captures activity across browsers and desktop applications, logging what users actually do rather than what they say they're doing. And a natural language policy engine allows security teams to write rules in plain English instead of wrestling with technical syntax.
In one deployment at an organization with roughly 5,000 employees, Jazz claims it reduced daily DLP noise from tens of thousands of low-confidence detections down to about 10 pre-investigated incidents—a reduction that, if replicable, would fundamentally change how security teams operate. The company detailed this case study in a blog post published March 6.
Early Traction, High Expectations
By the time Jazz came out of stealth, it had already signed 15 paying customers, according to Israeli business publication Calcalist. That's unusual velocity for an enterprise security product—categories where sales cycles typically stretch for months and pilot programs are the norm.
Named customers include Lemonade, the insurance upstart known for its tech-forward approach; AlphaSense, a market intelligence platform used by Wall Street analysts; and CAVA, the fast-casual restaurant chain that went public in 2023. It's a varied roster, spanning finance, insurance, and retail—exactly the kind of cross-industry validation that investors look for.
"Traditional DLP forces organizations to choose between security and productivity," Livneh said in statements accompanying the funding announcement. The framing is deliberate: security teams are tired of being seen as the department that slows everything down.
Kobi Samboursky of Glilot Capital Partners wrote in a March 10 blog post that securing more than a dozen paying customers in the first year represented "unusual traction" for this category. Perhaps. Or perhaps it signals that the old approach to DLP has finally worn out its welcome.
A Crowded, Complex Market
Jazz isn't alone in recognizing the opportunity. The broader DLP market has attracted attention from both incumbents and startups, particularly as generative AI tools create new exfiltration risks. Microsoft added AI-specific data controls to its Purview platform during 2025. Netskope, the cloud security giant, did the same. MIND, another startup, raised $11 million in October 2024 with a similar promise of smarter detection.
According to Verizon's 2025 Data Breach Investigations Report, roughly 60 percent of breaches involve some human element—the exact problem category that DLP systems are designed to address. That statistic suggests enormous market potential. It also suggests that existing solutions aren't working nearly as well as enterprises need them to.
The challenge for Jazz—and for any startup selling into this space—is execution at scale. Demonstrating that an AI agent can make intelligent judgments about data movement in a controlled environment is one thing. Proving it works across thousands of employees, dozens of SaaS applications, and the inevitable edge cases that emerge in real-world deployments? That's another matter entirely.
What $61 Million Buys
Jazz hasn't spelled out exactly how it plans to deploy the capital, though the job postings and early customer momentum offer clues. Scaling an enterprise security company requires significant investment in sales, customer success, and engineering—particularly when the product involves novel AI techniques that need ongoing refinement.
The company positions itself as rethinking DLP "for the AI era," language that appears throughout its website and recent blog posts. That framing is both marketing and product roadmap, suggesting that generative AI use cases—employees feeding sensitive data into ChatGPT, say, or inadvertently training external models with proprietary information—will remain a development focus.
Whether Jazz can deliver on its alert-reduction claims at scale, across industries and deployment scenarios, remains an open question. But $61 million provides considerable runway to figure it out. And in a market where the incumbent solutions have left enterprises frustrated enough to take bets on unproven startups, that might be enough.
For now, Jazz has what every early-stage company needs: capital, customers, and a thesis about why the old way of doing things is broken. Everything else is execution.
