The Medvi Story Is Either a Silicon Valley Miracle or a Warning Sign. Maybe Both.

Matthew Gallagher looked much less like the founder of a billion-dollar company than the person who might be working as a waiter at the Soho House club in West Hollywood on a recent afternoon. He had messy curly hair, a baggy T-shirt, and visible tattoos on his hands and arms. He revealed to the New York Times reporter that he had spent nearly every waking moment for eighteen months working on Medvi. In order to make personal appointments without interfering with his work, he even used AI to clone his own voice. He claimed to be earning over $3 million every day. He claimed to be lonely as well.

The Medvi story went viral in early April 2026 due to this combination of enormous financial success, structural isolation, and a story that is both deeply ambiguous and inspirational. At the same time, it began to draw more challenging questions than the initial coverage had posed.

The New York Times, which examined Medvi’s financial records, confirmed the facts, which are truly startling. With $20,000 and a number of AI tools, including ChatGPT, Claude, Grok for code and copy, Midjourney and Runway for images and video, and ElevenLabs for voice, Gallagher founded the telehealth business in September 2024. With no employees, he used these tools almost exclusively to create the website, create the ad creative, write the product descriptions, and set up the customer support

nfrastructure. In its first month, Medvi had 300 clients, and in its second, it added 1,000 more. In its first full year of operation, Medvi generated $401 million in sales and a 16.2% net profit margin by the end of 2025. These numbers are comparable to those of Hims & Hers Health, a publicly traded telehealth competitor that employs over 2,400 people and managed a 5.5% net margin during the same period.

Medvi’s product is GLP-1 weight-loss medication, specifically compounded forms of tirzepatide and semaglutide, the active ingredients in Wegovy and Mounjaro, which are filled by state-licensed pharmacies and prescribed via its partner platform CareValidate. There was a genuine and massive demand. Americans desired GLP-1 medications that were affordable and did not require the inconvenience of regular doctor visits. For a first month’s supply, Medvi charged as little as $179, pricing competitively while growing at a rate that rivals found hard to comprehend. The co-founder of CareValidate recounted speaking with Gallagher over the phone and inquiring as to whether he had an army of supporters.

This story seems to be being told in two registers at the same time. In one sense, it is a genuine example of what AI tools can accomplish when used by someone with true technical fluency and marketing instincts. Gallagher learned to code as a teenager by creating a fan page for Weird Al Yankovic, spent years switching between tech jobs and unsuccessful startups, and now has the perfect skill set at the perfect time. The CEO of OpenAI, Sam Altman, reportedly stated that he would like to meet Gallagher and that he seemed to have won a bet with his tech CEO friends. In 2024, Altman had predicted that a one-person billion-dollar company would eventually emerge.

The story is much messier in the other register. On February 20, 2026, the FDA sent a warning letter to MEDVi LLC, citing issues with the company’s compounded tirzepatide and semaglutide products, particularly the way they were marketed and labeled on the website. This raised concerns about whether the company was misrepresenting compounded medications as being on par with FDA-approved brand-name drugs. Approximately six weeks prior to the publication of the New York Times profile, that letter arrived, and critics believed the newspaper had undervalued its importance. The following month, Medvi was the target of a class action lawsuit alleging violations of California’s anti-spam law and that the company’s affiliate marketers were sending spam using spoof domains and falsified header information to avoid spam filters.

AI-generated before-and-after weight-loss photos with digitally altered faces, a scrolling ticker of media logos that made the company appear to have been featured in Bloomberg and the Times when it had only advertised there, and a customer service chatbot that occasionally invented drug prices—which Gallagher claims he honored—as well as false claims that the company sold hair-loss products before it actually did were all features of the early Medvi website, which Gallagher has acknowledged and partially corrected. Additionally, Medvi is a marketing and branding company that operates on top of platforms like CareValidate and OpenLoop, which manage the doctors, pharmacies, compliance, and shipping. In technical terms, Medvi is not a company with its own medical infrastructure. While that is a valid business model, it is not the same as “AI-built healthcare company.”

It’s possible that Medvi is precisely what the optimistic version of the story claims to be: an early, striking illustration of what competent individual operators can accomplish when AI tools reduce construction costs to nearly nothing. It’s also possible that the FDA letter and class action are the first indications of a more complex reckoning, and that the business model has regulatory vulnerabilities that haven’t fully emerged yet. It is possible for both to be true simultaneously. Gallagher appears to be aware of this, which could be the reason he was hesitant to share the story in public. He had good reason to be.