VR's Next Frontier Is... Visuals For Data Science?

VR's Next Frontier Is... Visuals For Data Science?
January 29, 2017
Virtualitics creates new data visualization techniques in virtual reality. The company has raised $3 million in seed funding. PHOTO:VIRTUALITICS


Virtual reality software isn’t all fun and games. 


Some startups are using it to help architects make realistic, virtual 3-D models, saving time and money that otherwise would be spent on physical models. Others are creating virtual workplaces, allowing workers to interact with multiple virtual computer screens when they’re really just using one computer and a virtual-reality headset.


On Thursday, Virtualitics LLC will launch its business that brings together data visualization and virtual reality. Its technology, which uses machine learning to summarize data patterns, is based on research done at the nearby California Institute of Technology and NASA’s Jet Propulsion Laboratory in Pasadena, Calif. Scientists there teamed up with video-gaming veterans who are helping Virtualitics develop user-friendly tools in virtual reality.


Virtualitics Chief Executive Michael Amori said virtual reality allows people to draw data insights that go well beyond a two-dimensional screen.


“When you’re looking at a financial-data set, there are many dimensions that matter,” Mr. Amori said. “With traditional software, you’re just looking at two things at a time.”


Unlike models that run on two axes in a spreadsheet, virtual reality gives users the opportunity to walk around and look at a data representation from multiple angles. Rather than looking at two data characteristics at a time, a viewer could compare eight different factors.


“You get a much better picture of what’s going on with the data,” he said.


Mr. Amori, who has more than a decade of experience in finance, shows off the product with exchange-traded funds. In trying to determine why some ETFs had better returns than others, Mr. Amori shows how Virtualitics can draw conclusions from a set of 30 different factors an analyst may have collected about different funds.


Pasadena-based Virtualitics has developed machine learning algorithms that help an analyst determine which factors will provide the best possible mapping. For example with ETFs, the machine picks volatility and expense ratio as the most important factors to consider.


The program then creates a multidimensional data representation. It uses natural-language processing to write a paragraph that helps interpret what the image means for people who don’t have data-science backgrounds.


Mr. Amori’s technology could be applied in many areas, from retail to urban planning—any sector that relies heavily on data analytics. The company said it has launched trials with several unidentified enterprise customers.


Virtualitics declined to identify all its backers by name but said a London-based financial institution, a South American retail group and several individuals invested in a $3 million seed round.


One investor, Andy Walter, a former chief information officer with Procter & Gamble, acts as an adviser. He said P&G had been an early adopter of virtual reality, and before he retired in September 2016, the company was using virtual reality to show retailers what shelf displays would like in stores.

Related articles

VRrOOm Wechat