Did you ever type worrying health symptoms into Google and panicked, because it said you’ve probably got cancer? Well, who didn’t. According to research, 3 out of 4 adults use Internet to search for medical advice. Now, there’s an alternative do Dr. Google: Infermedica, a diagnostic engine that uses Artificial Intelligence. Their core product – Infermedica API – was launched only in 2015, but it has already helped in prescreening over a million of patients.
Surprisingly, the main focus of Infermedica is not being the best diagnosis platform. “Unlike other companies, we are dedicated to being an Artificial Intelligence backend for other healthcare companies,” says Anna Szymczak, the COO at Infermedica. “We aspire to be a tool for others so third parties can build more powerful and intelligent products for patient and providers. Our focus is to provide a diagnostic engine which can be easily integrated into any frontend.”
Infermedica has three founders: a software engineer and entrepreneur Piotr Orzechowski, who is the current CEO, Dr. Roberto Sicconi – former IBM Watson Research Leader with significant experience in healthcare, and Dr. Irving Loh – a US-based doctor with over 30 years of clinical experience. The story of this startup began in 2012 with the creation of a symptom checker called Doktor-Medi.pl. Together with its English-language counterpart Symptomate.com, it provided advice to over a million patients thus far and is placed in top ranks of mobile symptom checkers in Google Play. As more and more health IT companies approached Infermedica to use their technology and database, this platform evolved into an AI diagnostic backend.
Infermedica has created an extensive database using multiple sources: from literature, published medical cases and reliable publications, through statistics (millions of anonymized patients records and common patterns in them), to a board of expert physicians acting as curators for the platform. Machine learning techniques enable constant improvement of the AI behind the diagnosis process. The platform acts like a real doctor, allowing the patient to describe his symptoms, asking additional questions and recommending next actions such as visiting an appropriate physician.
Since the launch in 2015, Infermedica API – a product for developers who want to start their own diagnosis engine – has grown from 0 to over 650 developers. It supports platforms such as Affidea, Neuca, Diagnostyka, HealthLoop or Gyant – a diagnostic chat bot. It also rolled out to several hospitals in the US.
The main sources of funding are three Polish-based investors – RTA Ventures, Innovation Nest and Venture Incubator – and licensing fees for using the Infermedica API. The fees are lower for clients who provide a valuable input to the medical knowledgebase.