2 years ago I was working as a software engineer at a large corporate. I was part of a team building a product for virtual classrooms and it inherently involved sensitive pupil data. I remember the day where I got the simplest request from our lead product: “I want to know the average grade of students in the classrooms to assess the impact of the system on exam performance”. Sounds like a legitimate 2-seconds analysis, right? wrong.
That’s when I first discovered the complexities of analyzing production data. An average exam grade could potentially be traced back to a specific student given additional information or queries. Statistics can either tell a broad story, or a very precise and targeted story, depending on the data, the questions and intent. Instead of extracting a simple average, I was only allowed to calculate approximate and very high-level figures.
It sounds cliché, but I swear I told myself “there has to be a better way”.
And there was.
During that time, Shachar was finalizing his Ph.D. in computer science, specializing in Differential Privacy. A mathematically quantifiable definition of privacy that was formulated precisely to unlock analytics without compromising individual privacy. We realized that this innovative approach has to be productized and available for medium to large companies, and decided to found PVML.
Value Proposition and Raising Our First Round
As a solid and completely non-committal first step, we both quit our day jobs immediately. Being a married couple and putting our household’s income on hold to start a company was definitely a risk, but we had classic founders’ optimism clouding our (typically numbers-led) judgment and a true confidence that despite infamous startup statistics – this cannot fail. Plus, an all-consuming 24/7 obsession was our secret weapon; being a couple for fourteen years turned out to be quite an unfair advantage given that work-life balance is a huge struggle for most startup founders.
Statistics turned out to be even worse that year; 2023 was a particularly difficult year for Israeli high-tech, with a 50% decrease in the number of companies who raised a Seed round compared to 2022. Despite that, we managed to demonstrate that our solution, a new generation of data access platforms that does not require a binary approach of either “see raw data” or “see no data”, is exactly what the market was missing. This was especially powerful in the midst of the AI revolution, where we have AI increasingly accessing sensitive data which raises the stakes. The old-school data anonymization methods are officially inadequate, while companies are actively looking to integrate AI and speed up time to insight using free text. With PVML we offered just that: democratizing data access in a way that facilitates AI, thanks to powerful Privacy-Enhancing Technology backing your data protection.
Data Peace Of Mind
PVML provides a secure foundation that allows you to push the boundaries.
In order to demonstrate the strength of our value proposition we decided to start our Seed fundraising by talking to a lot of potential clients rather than investors. Our excitement reached new highs every time we heard the pain companies described regarding data access and even managed to generate leads that we later converted to PoCs and clients.
We closed our $8M Seed round during the second half of 2023, led by elite VC NFX (with General Partner Gigi Levy-Weiss). We earned the confidence and backing of Gefen (alongside Managing Parter Limor Ganot) and FJ-Labs (with Partner Jeff Weinstein), as well as key figures in the data security industry such as Ran Nahmias, Amit On, Erez Yarkoni and Jonathan Lebowitsch. Their support played a pivotal role in the inception of PVML.
The Dream Team
We believe that people are the most important factor in the success of any goal that requires collaboration, innovation and courage. Our goal definitely required all three and more, so we knew we needed no less than a super-star team of engineers and researchers by our side. We reached out to our most talented former colleagues and friends from university, inviting them to join our crazy journey. The response was overwhelming – leading technological minds from giant corporates to fast-growing startups came onboard. Each bringing their unique talents and strengths, we’ve assembled an exceptional team, poised to pave the way for groundbreaking accomplishments at PVML.
Introducing PVML’s Data Access Platform
To enable seamless integration, we first enable users to connect any database using a simple connection string, making it a 2-minute process that requires no infrastructure changes. Admins get one single point to monitor and control all their data access, get a high level view over queries, permissions, user requests and to-dos. We track excessive permissions, unused permissions or conflicting permissions to help reduce managerial overhead with 1-click actions that boost security.
We built an SQL compiler able to translate database queries to Differential Privacy queries on the fly and without the user needing to alter the query syntax. This guarantees privacy and permissions enforcement for any way the users choose to interact with data, whether it be SQL, AI or through an external BI tool – PVML is in charge of live output-level data protection.
