The parties involved in the so-called "venture lab" include Andreessen Horowitz, The Social + Capital Partnership and Anand Rajaraman, a computer science professor at Stanford who founded Kosmix, the company that became WalmartLabs, the tech-development division at Walmart.
According to its backers, Data Elite will fill a need for big data startups that have found generic startup incubators and investment funds unsatisfactory, since the people involved all too often lack an understanding of the specific challenges and possibilities of data science.
Data Elite's three-month incubator program will provide a minimum of $150,000 to each startup through common stock purchase and convertible notes. The inaugural class will include up to 10 companies, all of whom will be expected to work at the Data Elite offices in San Francisco.
"We would like to create a big data community, a big data hub," says Stamos Venios, co-founder and managing partner of the fund. "We want them to stay in and keep working until they outgrow the space." To that end, he says the venture lab will provide a total of 10,000 square feet for the companies to use.
Along with investment capital and workspace, the main draw will be Data Elite's team of 12 data-science experts, which includes Ken Rudin, Facebook's head of analytics; Jeff Magnusson, manager of data science platform architecture at Netflix; and Rajaraman. All of them have signed agreements saying they will be present at Data Elite's offices at least five hours a week to coach the startups.
"The hardest part wasn't the fundraising, it was getting these guys on board," says Venios. "They aren't doing it for the money. They want to see the next wave of data scientists."
In addition, representatives of Fortune 500 companies will stop by to advise Data Elite startups on problems that big data could solve. The startups will be able to begin tailoring their work to these potential customers early on.
As its name suggests, Data Elite won't bring on just any startup. To be considered for the program, a startup's founder must have at least five years of experience in data science, or a successful exit from a prior big data company. "We want really, really high-profile engineers to come to us for funding," Venios says.
Although the fund will focus on early-stage startups, it is currently interpreting that mandate somewhat loosely. One company with which Venios has spoken about entering the program has already raised more than $2 million in a series A round, but it wants to pivot and its founder thinks Data Elite's expertise could be helpful.
Venios says Data Elite has not signed any contracts yet. Startups have until December 15 to apply, and the program will start on January 15, 2014.