Navigating the scores of wines available on a trip to Napa Valley six years ago, Houston entrepreneur Amy Gross wondered why there wasn’t a more scientific way to pick the perfect glass.
“If music can be characterized objectively like Pandora, why can’t wine?” says Gross, recalling a conversation with her husband. “Why do we have to listen to others to find out what we like?”
That question led her to launch Wine4Me, and today, along with several other Texas startups, it is among the latest batch of companies to join IBM Watson’s Developer Cloud. Wine4Me uses artificial-intelligence techniques to recommend wines based on a consumer’s preferences. The company says that kiosks, with an iPad interface, will be rolled out in grocery stores in early 2016.
“The consumer can touch the screen, looking for a fruity wine that goes with salmon and is less than $15,” Gross says. “It then returns results of what the grocery store has.”
Wine4Me’s parent company, VineSleuth, joins more than 25 other organizations innovating in healthcare, recruitment, financial services, and e-governance. The announcement was one of several by IBM Watson Thursday, which also said that Big Blue was opening a second Watson headquarters in San Francisco to operate in tandem with its office in New York.
In Texas, Watson is being put to use at the University of Texas M.D. Anderson Cancer Center in Houston to develop the Oncology Expert Advisor tool, which is designed to leverage “cognitive computing” capabilities to research the best treatment options for cancer patients.
And Manoj Saxena, who formerly led Watson’s operations in Austin, hopes to encourage a cognitive computing startup cluster through investments made by the Austin branch of The Entrepreneurs’ Fund, established last year.
For Gross, who founded VineSleuth in 2009, the Watson partnership is welcome validation for her startup, which began as a random thought during a visit to Napa. “Working with such an established company, partnering with Watson, this opens doors to us to reach new clients that may not have opened before,” she says.
She’s already planning further development of Wine4Me, including creating white-label versions of the app for grocery chains. “If you are looking for a specific wine that’s not at the store, it will show other wines that are like that that are at the store,” Gross says.
Down the line, she adds, all those customer searches are rich data analytics opportunities for retailers. “We track all the wines people enter, share which ones people are requesting [that] they don’t carry,” Gross says. “We’ll keep track of that for them.”
And people don’t have to use “wine-speak” to find that perfect bottle, she adds. “The average bottle of wine that’s sold is less than $10,” she says. “People are buying a lot of grocery store wine. We make sure they get what they really want.”
The other Texas companies in the Watson program are:
—WayBlazer (Austin): Uses a natural language interface to act as a travel concierge. The company’s “discovery engine” uses Watson technology to absorb data to link places, offers, and preferences, with social, cultural, and economic data tagged to recommendations for each consumer. The company is led by Terry Jones, the founder of Travelocity and a founding chairman of Kayak.
—Bloomfire (Austin): Cloud-based software that helps employees within a company easily find the information they need to do their jobs. By scanning posts within the software and automatically creating tags via the Watson data insight APIs (application programming interfaces), employees at companies such as Whole Foods could find company procedures and guidelines pertinent to specific tasks quicker.
—Cerebri (Austin): These student entrepreneurs were winners in the Watson University Competition in January. Cerebri works with the United Way for Greater Austin to help deliver social services information via mobile apps, especially for people without home Internet service. Cerebri gathers information about each app user, which is then analyzed by Watson to come up with a list of services that will most benefit the user.
