If you’ve ever woken up feeling under the weather and gone online to check your symptoms, you’re not alone — 35 percent of adults in the U.S. turn to the internet to self-diagnose medical conditions, according to the Pew Research Center’s Internet & American Life Project. The trouble is, internet diagnoses tend to be less than spot-on. In fact, online health tools only provide accurate results about 34 percent of the time.

That got Allon Bloch, the former CEO of Wix and Vroom, thinking: What if instead of spitting out generic lists of ailments in response to queries, online symptom checkers personalized results according to searchers’ health histories?

Enter K Health, a Tel Aviv, Israel-based primary care startup that today announced a $12.5 million funding round. Participating investors include Mangrove Capital Partners, Lerer Hippeau Ventures, Primary Ventures Partners, the Box Group, Max Ventures, Bessemer Ventures, and Comcast Ventures. The company is launching in New York City following a trial run in Madison County, Alabama earlier this year.

“Today, if you go online and search for something as simple as a cough, you’ll see millions of results, ranging from the common cold to cancer,” Bloch said. “With K, we use real data from millions of people so you discover and understand the medical outcomes of people like you and have informed conversations with your providers about treatment options.”

K Health, unlike most online self-diagnostic platforms, takes into account detailed medical histories, clinical outcomes of millions of people, and the experience of thousands of doctors to deliver highly relevant treatment information on hundreds of diseases. Thanks to partnerships with local health systems, it also offers same-day appointment bookings with physicians, as well as complimentary remote consultations.

“We … help people understand what they might have based on this … comparison. It emulates the best doctor in the world,” Bloch said.

K Health sources from a database of millions of electronic health records and 2 billion “health events” — maladies like nausea, headaches, and vomiting — supplied by Maccabi, Israel’s second-largest health fund. Users start by downloading an app for iOS or Android and answering roughly 20 questions about their age, gender, body mass index, health history, and symptoms. A machine learning-powered backend uses the responses to build a private profile, which it compares to medical insights gleaned from physician notes and charts.

The results page shows a list of outcomes experienced by people in similar health circumstances, along with a percentage indicating the likelihood of each diagnosis. Reports and profiles can be shared with clinicians via a HIPPA-compliant messaging feature ahead of telemedical or in-person appointments, if users so choose.

“Everything we do is driven by a relevant comparison,” Bloch said. “When you go online and something’s bothering you … it’s a scary moment. It’s like a broken clock that’s right twice a day. That’s why we [take this] very accurate approach to understanding potential diagnoses.”

Bloch said that K Health has already improved patient outcomes in markets where it has launched. Integrity Family Care, a Madison County health care provider that was one of K Health’s first stateside partners, says that in two months it was able to manage care for 85 percent of patients using the app remotely, outside of emergency rooms and offices.

“There’s so much misdiagnosis — so much overtesting and overtreatment,” Bloch said. “We see [K Health] as a tool that can help to rule out something that’s potentially serious [and] help users take control of the information … [For] people who don’t have the ability to get access to quick care, that’s crucial.”

K Health’s accuracy aside, there’s growing concern among health care professionals that symptom checkers needlessly fuel patient anxiety and confusion, a phenomenon known as “cyberchondria” (a portmanteau of “cyberspace” and “hypochondria”). A 2013 study conducted by Baylor University found that people with a high intolerance for the unknown tended to handle self-diagnostic results poorly.

“If I’m someone who doesn’t like uncertainty, I may become more anxious, search further, monitor my body more, go to the doctor more frequently — and the more you search, the more you consider the possibilities,” Thomas Fergus, a lead author on the Baylor study, said in a statement. “If I see a site about traumatic brain injuries and have difficulties tolerating uncertainty, I might be more likely to worry that’s the cause of the bump on my head.”

Bloch noted that K Health makes it clear its results aren’t meant to replace the advice of medical doctors, and he added that the team is experimenting with ways to ensure treatment suggestions aren’t misinterpreted.

“We agonize over [these point[s],” he said. “We’re focused on making sure people understand [the results].”

K Health is funded in part by Maccabi and Morris Kahn Institute for Research and Innovation, the tech incubation arm of Tel Aviv, Israel-based health maintenance organization (HMO) Maccabi Health. It says that any profits from the app will be reinvested into K Health’s database.

Sign up for Funding Daily: Get the latest news in your inbox every weekday.