The Story of PVML: The Company Building the Future of Data Enablement

A product manager’s simple request for average exam statistics exposed a fundamental flaw in how the world’s most powerful tech companies handle sensitive data. In a recent episode of Category Visionaries, Rina Galperin, CTO and Co-Founder of PVML, a data access platform that’s raised $8 million, shared how that moment of frustration at Microsoft became the genesis of a company redefining data security.

The Microsoft Moment

Two years before founding PVML, Rina was a software engineer at Microsoft, working on a virtual classroom solution. The product involved exactly the kind of sensitive data you’d expect: student information, exam tasks, performance reports. Then came what should have been a straightforward request.

Her product manager needed basic statistics—an average exam grade in a specific region to measure the impact of their system on exam performance. “It sounds like a legitimate two second analysis, right? Well, turns out not really,” Rina recalls. “I mean, this was the first time ever I actually encountered the complexities of analyzing production data.”

The problem wasn’t technical capability. Microsoft had the data, the infrastructure, the engineering talent. The problem was privacy. An average exam grade, combined with other available information or queries, could potentially be traced back to individual students.

“An average exam grade could potentially be traced back to a specific student given additional information or queries. So it turns out statistics are tricky,” Rina explains. “They can either tell a broad story, a very precise story, or even a malicious story. It all depends on the data, the question, and the user’s intent.”

Instead of the simple analysis her PM requested, Rina could only calculate approximate, high-level figures. The data was there. The business need was legitimate. But the tools to safely bridge that gap didn’t exist.

Her reaction carries the frustration every founder feels when they stumble onto their company’s origin story: “As cliche as it sounds, I swear I told myself there has to be a better way. I mean, if one of the most powerful tech corporates is struggling with sensitive data access, then everybody else is just doomed.”

When PhD Research Meets Real-World Pain

The timing of Rina’s frustration proved fortunate. While she grappled with production data limitations at Microsoft, her future co-founder Shachar was completing his PhD in computer science, specializing in something called differential privacy.

“A mathematically quantifiable definition of privacy. And yeah, that’s a thing. It actually exists,” Rina notes with the slight amazement of someone who discovered an elegant solution to an impossible problem.

Differential privacy was formulated precisely to solve the challenge Rina faced: unlocking analytics without compromising individual privacy. “This definition was formulated precisely to help you unlock analytics without compromising privacy. So that’s kind of how problem met solution. And it just clicked and we decided to found PVML.”

The elegance of this founding story matters. PVML wasn’t born from a whiteboard brainstorming session about market opportunities. It emerged from the collision of genuine pain at one of the world’s most sophisticated tech companies and cutting-edge academic research that most practitioners didn’t know existed.

Door Knocking Without a Network

Armed with a compelling problem and an elegant solution, Rina and Shachar faced a new challenge: they had no idea how to sell enterprise software.

“We both were quite new to the whole entrepreneurial world, first timers, two techie people, and we really just went door knocking,” Rina shares. “We had close to zero inside connections, and we’re new to the startup ecosystem. We just started reaching out to people on LinkedIn and email and just trying to get meetings booked.”

Landing their first paying customer took approximately six months. For technical founders accustomed to the immediate feedback loops of engineering, six months of rejection and iteration feels eternal. But that period proved foundational.

Rather than obsessing over product features or pricing models, they focused on understanding people. “We spent a lot of time digging into the problem domain and understanding the pain of not only our potential clients, as in the companies, but specifically the buyer in the company, what drives and motivates him, what sets him back, and so on,” Rina explains.

This buyer-level focus—understanding individual motivations, career concerns, and decision-making constraints—became PVML’s early go-to-market edge. “So it really helped us target our ideal customer profile very well and lend our first clients.”

Redefining a Restrictive Category

PVML entered the data access market facing a positioning challenge. Every existing solution in the space focused on restriction: anonymization, masking, encryption. The underlying philosophy was simple and limiting.

“Basically, let me hide or remove your data because I don’t want to risk a data breach,” Rina summarizes. These approaches made sense given available technology, but they created a fundamental tension. Organizations needed to analyze data to drive business value, but security solutions forced them to restrict, limit, and hide that very data.

PVML’s differential privacy technology flipped this paradigm. Instead of choosing between security and utility, organizations could have both. “The first technology that actually helps unlock more data for analysis without altering it and without risking it,” Rina states.

But differential privacy alone wasn’t enough. PVML layered in a second innovation: agnostic data access. Most organizations face a frustrating reality—implementing permissions in one tool, then facing the choice of duplicating that work when adding new tools or simply avoiding innovation entirely.

“Think about a company that implemented information management in, let’s say, power BI. So a year later, the analysts come and they want to use tableau. So now this company is quite stuck because it can either implement a whole new set of access rules in tableau and then double its admin overload by having to manage two sources of permissions, or the company will simply avoid adding this new bI tool to its toolset, which is also a bad option.”

PVML’s solution centralizes access to all data sources without moving or duplicating anything, enabling permission management across all access interfaces—BI tools, APIs, and now AI. “I would say it’s data access, but it’s enabling instead of limiting.”

Creative Marketing in a Crowded Space

With cybersecurity being notoriously saturated, PVML adopted an unconventional marketing philosophy. “I think it’s important to be bold and take risks with marketing. Otherwise you’re just blending in,” Rina states.

Their tactics range from simple to surprisingly personal: coffee vouchers handed out at conferences to land prospect meetings, customized mugs created for client onboarding days. “You sometimes need to risk having a door shut at your face to potentially land something big and suppress competition.”

Perhaps more strategically, PVML stopped prioritizing cybersecurity-specific events. Instead, they targeted domain-specific conferences in verticals showing traction: fintech, mobility, healthcare. “We do try to look for clients not specifically in, I mean, you mentioned events and not specifically in cyber events, but maybe branch out more to domain specific events that we’re seeing traction specifically in fintech, mobility, healthcare.”

The Future: Conquering Markets and Verticals

PVML’s three-to-five-year vision centers on geographic expansion and vertical domination. “We want to conquer the european and american markets, especially in areas where we’re market leaders, such as mobility, transportation, but also healthcare and fintech, where we see significant traction,” Rina shares.

Currently operating between Tel Aviv and New York, the company is actively expanding into the US market. The bet is straightforward: the ground-level demand they’re seeing will translate into substantial American business growth.

The broader vision extends beyond geography. As organizations increasingly face the tension between data protection and data utilization—a tension only intensifying with AI adoption—PVML’s enabling approach to data security positions them at the center of a fundamental shift in how companies think about sensitive data.

What started as a frustrated Microsoft engineer unable to calculate simple statistics has evolved into a company reimagining data security not as restriction, but as enablement. For organizations drowning in compliance requirements while desperate to unlock insights from their data, that shift can’t come soon enough.