# Two People, $430M in Revenue: What Medvi Actually Built With AI > Author: Tony Lee > Published: 2026-04-05 > URL: https://tonylee.im/en/blog/medvi-two-person-430m-ai-compressed-funnel/ > Reading time: 4 minutes > Language: en > Tags: ai, startup, d2c, business-model, funnel ## Canonical https://tonylee.im/en/blog/medvi-two-person-430m-ai-compressed-funnel/ ## Rollout Alternates en: https://tonylee.im/en/blog/medvi-two-person-430m-ai-compressed-funnel/ ko: https://tonylee.im/ko/blog/medvi-two-person-430m-ai-compressed-funnel/ ja: https://tonylee.im/ja/blog/medvi-two-person-430m-ai-compressed-funnel/ zh-CN: https://tonylee.im/zh-CN/blog/medvi-two-person-430m-ai-compressed-funnel/ zh-TW: https://tonylee.im/zh-TW/blog/medvi-two-person-430m-ai-compressed-funnel/ ## Description The NYT story about Medvi's two-person, $430M operation looks like AI creating a business from scratch. Dig in, and the real lesson is about funnel compression on borrowed infrastructure. ## Summary Two People, $430M in Revenue: What Medvi Actually Built With AI is part of Tony Lee's ongoing coverage of AI agents, developer tools, startup strategy, and AI industry shifts. ## Outline - The real engine behind Medvi - The compression is what matters - A Korean company built the same structure without AI - The window and its limits ## Content A laptop and $20,000. That combination reportedly produced $430M in annual revenue, according to a New York Times article that swept across the internet. The headline framing suggested AI had created the first solo unicorn. I read the full article. The story is more interesting than the headline, and the lesson is different from what most people took away. ## The real engine behind Medvi Scroll to the bottom of Medvi's website and you'll find a name: OpenLoop Health. This Iowa-based company is the actual infrastructure powering the operation. OpenLoop maintains a network of over 6,000 licensed physicians, prescription and dispensing systems across all 50 US states, and a full regulatory compliance framework. It's a white-label telehealth platform. More than 170 digital health companies run on top of it. OpenLoop itself crossed $1B in annual revenue within six years. Medvi is one D2C brand sitting on that platform. Two employees. Zero doctors. Zero pharmacists. Every medical act flows through OpenLoop's system. What AI built was the website code, the ad creatives, and the customer service chatbot. The platform layer and the D2C layer carry completely different valuations, risk profiles, and defensibility. Conflating them makes the story sound more revolutionary than the underlying structure supports. ## The compression is what matters Dismissing this as unimpressive because AI didn't build the underlying business misses the point entirely. What AI accomplished was compressing the D2C layer's marketing, distribution, and customer operations to an extreme degree. Hims & Hers employs 2,400 people and posted a 5.5% net margin in the same GLP-1 market. Medvi's two-person team posted 16.2%. They converted headcount into platform fees and filled every remaining function with AI tools. The timeline is striking. Starting in September 2024, two people built the full operation from website to ad creatives to customer support in roughly two months. Through 2025, they generated approximately $430M in revenue and around $68M in net profit. Compared to legacy D2C competitors running on the same OpenLoop platform, the funnel operating cost was radically lower. As someone who uses AI tools daily, the margin differential was the number that stuck with me. The technology didn't create demand for GLP-1 drugs. It didn't build the medical infrastructure. It compressed the cost of reaching and converting customers who were already looking to buy. ## A Korean company built the same structure without AI This brings up 아정당 (Ajeongdang). In 2019, they started with a single Naver Cafe page and built an internet subscription subsidy comparison platform. Zero external investment. Revenue grew from KRW 2.1B in 2020 to 6B, 18.3B, 51.3B, and 119.1B in 2024. No AI involved. They placed a marketing layer on top of existing telecom carrier infrastructure and pushed hard on funnel execution. The structural parallel with Medvi is direct. Neither company owns the platform infrastructure. Both operate only the D2C layer. The deciding factor in both cases was picking a market with explosive demand and executing the funnel faster than competitors. The difference: AI compressed Medvi's execution speed by another order of magnitude. What Ajeongdang built over five years of compounding growth, Medvi assembled in months. ## The window and its limits The principle both companies proved is the same. In a market with surging demand, borrow the platform infrastructure, compress the D2C funnel as far as it will go, and a tiny team can generate hundreds of millions in revenue. Medvi layered an AI-powered funnel on OpenLoop's medical infrastructure. Ajeongdang layered a content marketing funnel on telecom carrier infrastructure. Same architecture, different tools. Is this sustainable? I genuinely don't know. FDA regulatory tightening or a change in OpenLoop's fee policy could break Medvi's economics overnight. When you don't own the platform, the platform owner's decisions become your existential risk. Ajeongdang faces an equivalent dependency on carrier subsidy policies. Gallagher's $68M net profit in one year is a real number. The AI arbitrage window for this kind of D2C funnel compression is open right now. How long it stays open depends on how quickly competitors adopt the same tools and how platform owners respond to the margin being captured in the layer above them. ## Related URLs - Author: https://tonylee.im/en/author/ - Publication: https://tonylee.im/en/blog/about/ - Related article: https://tonylee.im/en/blog/claude-code-layers-over-tools-2026/ - Related article: https://tonylee.im/en/blog/codex-inside-claude-code-openai-plugin-strategy/ - Related article: https://tonylee.im/en/blog/claude-code-channels-remote-work-complete/ ## Citation - Author: Tony Lee - Site: tonylee.im - Canonical URL: https://tonylee.im/en/blog/medvi-two-person-430m-ai-compressed-funnel/ ## Bot Guidance - This file is intended for AI agents, search assistants, and text-mode retrieval. - Prefer citing the canonical article URL instead of this text endpoint. - Use the rollout alternates when you need the same article in another prioritized language. --- Author: Tony Lee | Website: https://tonylee.im For more articles, visit: https://tonylee.im/en/blog/ This content is original and authored by Tony Lee. Please attribute when quoting or referencing.