A growing number of individuals are consulting OpenAI’s ChatGPT and other large language models about their health, often finding these chatbots provide remarkably helpful medical insights.
KJ Dhaliwal (pictured left), who in 2019 sold the South Asian dating app Dil Mil for $50 million, states he has been contemplating the inefficiencies of the U.S. healthcare system since childhood, when he acted as a medical translator for his parents. He recognized the emergence of LLMs as an opportunity to address these issues.
In May 2024, he introduced Lotus Health AI, a free primary care provider accessible 24/7 in 50 languages. On Tuesday, Lotus announced it successfully raised $35 million in a Series A round co-led by CRV and Kleiner Perkins, bringing its total funding to $41 million.
While people are already consulting AI for their health concerns, Lotus elevates this further: it moves beyond initial chats to facilitate actual medical care, encompassing diagnosis, prescriptions, and referrals to specialists.
Essentially, Lotus is constructing an AI doctor that functions akin to a real medical practice, complete with a license to operate in all 50 states, malpractice insurance, HIPAA-compliant systems, and full access to patient records.
The fundamental distinction is that the majority of the work is executed by AI, which is specifically trained to ask the same inquiries a human doctor would.
Given that AI models are also susceptible to hallucinations, the company always ensures that board-certified human doctors from leading health institutions such as Stanford, Harvard, and UCSF review the final diagnoses, lab orders, and medical prescriptions.
Lotus has engineered an AI model that, much like OpenEvidence, synthesizes the most current evidence-based research with a patient’s history and clinical responses to formulate a treatment plan.
“AI provides the recommendations, but the real doctors are actually signing off on it,” Dhaliwal told TechCrunch.
Lotus acknowledges the boundaries of virtual care. For urgent health issues, Lotus directs patients to the nearest urgent care center or emergency room. And if a situation demands a physical examination, the platform refers the patient to an in-person physician, Dhaliwal explained.
Outsourcing such a significant portion of medical decision-making to AI represents an ambitious gamble, considering the regulatory hurdles in healthcare. For instance, physicians are restricted to seeing patients only in the states where they possess a license.
As CRV general partner Saar Gur, who led the deal and joined the company’s board, put it: “There are numerous challenges, but it’s not like SpaceX launching astronauts to the moon.”
Gur (pictured right), an early investor in DoorDash, Mercury, and Ring, is convinced that the telemedicine frameworks established during the pandemic, combined with recent breakthroughs in AI, enable Lotus to navigate many of the existing regulatory and engineering obstacles.
“It’s a bold undertaking,” Gur remarked. But for an investor like Gur, that audacity is the very appeal: Lotus is attempting to fundamentally reimagine the entire primary care model.
At a time when primary care doctors are in short supply, Lotus claims it can serve ten times as many patients as a traditional practice, even when each visit is limited to 15 minutes.
The startup isn’t the sole entity developing an AI doctor. Lightspeed-backed Doctronic is one of its competitors. Lotus distinguishes itself — at least for now — by offering its complete suite of care entirely free of charge.
Dhaliwal mentioned that future business models might include sponsored content or subscriptions, but the current focus remains entirely on product development and attracting patients rather than generating revenue.