Explorium Nets $75 Million to Grow Alternative Data Platform
When your data science initiative hits a wall with commoditized machine learning algorithms, the best solution is often to seek new data, according to Maor Shlomo, the CEO of Explorium, the alternative data platform provider that today announced it has raised $75 million in a Series C round.
“You get to the point where predictive models that used to be very complex and sophisticated are now easier to do,” Shlomo told Datanami. “When the challenge is not any more how to train algorithm A or B or how to spin up a cluster to analyze a lot of data, the challenge is back to what is the right data to feed into those algorithms to make them as robust, as accurate, as predictive as we want them to be.”
That’s where Explorium comes in. The company ostensibly runs a marketplace for alternative data that falls into one of five buckets: data about companies, people, locations, products, and time-series data.
But on top of this data marketplace, Explorium adds its own data science twist. The company uses predictive capabilities to help customers figure out which data sets have the best chances of impacting its customers’ businesses. And once the right external data set with the right variables has been tested and vetted, Explorium helps the customer integrate that data with their own processes.
An engagement with the San Mateo, California, company often starts with a customer sharing the types of data sources they already have, Shlomo says. Explorium can run this internal data source through its own predictive algorithms to identify other external data sets that could potentially help the customer.
“So basically, it’s looking for correlations across tens of thousands or hundreds of thousands of potentially relevant variables, and allowing the customer to identify the top 1% of variables that can actually be impactful for the question they have,” Shlomo explains. “We bring in data about leads, about businesses, about people, customers, locations and stuff like that, then you use the platform as a way to do automated matching and joining of data and discovery of correlations and variables.”
Explorium’s platform includes commonly used machine learning algorithms and environments, including scikit learn, XGBoost, Tensorflow, and others. The company uses those algorithms behind the scenes to automatically rank which data sources are most applicable to customers’ specific needs, whether its fraud detection or propensity to buy or some other data science challenge.
Many organizations today are struggling not only to differentiate on the data science and engineering side of the equation, but they’re also facing friction with internal business processes when trying to bring new data sources online, Shlomo says.
“If data is the new oil, then you’re constantly searching for more oil sources,” Shlomo says. “But the process now in a lot of enterprise is so painful and so rigorous, you have to involve procurement and legal and data science to test the data, and privacy, and engineering for integration – all for one data source that might turn out to be not so valuable.”
The company, which was founded in 2017, occupies a unique spot in the data and analytics world, according to Shlomo. “We’re not a data vendor. We’re not an automated machine learning player,” he says. “We’re sitting on the fence in between bringing the world of external data into analytics process that analysts and data scientist are caught behind.”
Explorium’s $75-million Series C was led by Insight Partners, with existing investors Zeev Ventures, Emerge, F2 Venture Capital, 01 Advisors and Dynamic Loop Capital participating. The round comes less than 12 months after its series B round, and brings the company’s total investment raised to more than $127 million.
The company, which also has offices in Tel Aviv, Israel, also announced that George Mathew has joined the board of directors. Mathew is a managing director at Insight Partners and was formerly president and COO of the analytics firm Alteryx.
The End of the Data Broker Model and the Rise of the Data Marketplace
Explorium Platform Billed as App Store for Predictive Models