Exploring Data Science With Explorium’s Or Tamir
“My career has been around startups all the way,” says Or Tamir, co-founder and COO of Explorium. ““Growing companies that make an impact is something I’ve always been passionate about.”
Because startups have composed the foundation and the scaffolding of Tamir’s career, he views the fact he ended up co-founding and helming a project like Explorium as almost inevitable. In a way, it was a logical next step: “founder” was the only job role in the world of startups that Tamir hadn’t yet tackled.
Or Tamir started on the investor side, working with venture capital and proceeding to business and product roles. He then worked extensively in early stage and growth companies—the word “mentor” appears frequently on Tamir’s CV—including leadership roles in one that grew from 30 to 300 people within a handful of years.
After all of that? “It’s been only natural for me to go on and be a full-on co-founder,” Tamir asserts. His experience, especially the bird’s-eye view of how startups operate gleaned from his years working with venture capital, has been profoundly useful—although, he says, he had to unlearn the skepticism that comes with venture capital work and focus instead on seeing the (many) positives.
And thanks to Tamir and his fellow co-founders—one is ex-Israeli military intelligence, one was previously a veteran of the ad-tech industry, and both have extensive experience in what Tamir calls “the data science side”—Explorium was born in 2017. The inspiration for the platform came in large part from the diversity of the co-founders’ backgrounds: they realized that they created such a strong team precisely because of the radically different perspectives they brought to the table, and got to thinking about how to translate the phenomenon of “stronger together” to data science.
“We all kind of came around the same basic problem, the same basic paradigm, which is that whenever you’re doing anything with data, be it simple BI analysis to advanced predictive modeling, you’re always limiting yourself to a small subset from the very beginning,” Tamir explains. “You go to the data you know, and you usually try the ideas you’re already familiar with, and you’re missing so much potentially from the picture.”
That’s where Explorium comes in. The focus of the automated platform is to lessen the burden of prediction, to help shoulder the time-consuming and labor-intensive load of finding relevant data and variables. Clients’ data sets are brought onto the platform, and the Explorium system, guided by customer prediction target, automatically finds new data sources and the right way to utilize them. Ultimately, Tamir says, Explorium “boils it all down to the ones that are very helpful and have the most predictive power and trains the model, that you can then consume, and we take it all the way down to production.”
These clients are generally mid-market data scientist–driven companies whose businesses orbit around predictive models. Many of these clients are already on AWS, which helped influence the “cloud agnositic” Explorium team.
As Explorium enters its second year, Tamir and his team are already thinking about the future. Product road maps are being finalized, and plans for 2020 and beyond are already being drafted. The primary focus of Explorium is the creation of what Tamir terms “a trove of data sources.” They’re also working on making the platform more extendable, explaining that Explorium customers enjoy coding and implementing their own ideas. This extension would allow Explorium and Explorium clients to work in even closer tandem.
The business side—Tamir’s previous expertise niche—is growing as well, straining new verticals, geographies, and new use cases. But Tamir has shed enough of his venture capital past to keep his vision for the future laser-focused, not to mention wholly optimistic. “A main focus for us is always the data itself,” he says. “Our goal is basically to make everything that’s online available to our customers.”