Luna Diabetes

It's 3 a.m., and the insulin is dosing itself

Somewhere in San Diego, a person with type 1 diabetes is asleep. They did not set an alarm to check their blood sugar. They did not stack a correction dose and pray. On their abdomen, a patch the size of a coin is quietly doing the math - reading glucose, deciding, delivering a micro-dose of rapid-acting insulin, and doing it again a few minutes later. By morning, more time was spent in range, and almost none of it was spent thinking about diabetes. That is the entire pitch of Luna Diabetes, and it is a strangely radical one.

Luna Health, Inc. - the legal name behind the friendlier brand - is a 16-person medical device company that has made a deliberate, slightly contrarian choice. While most of the industry races to build better full-time insulin pumps, Luna built a pump for the people who looked at full-time pumps and said, politely, no thank you.

Automation arrived. Most people were left out.

Automated insulin delivery is one of the genuine success stories of modern diabetes care. Pair a continuous glucose monitor with a pump and a clever algorithm, and you can hand off a huge amount of the relentless, every-few-hours arithmetic that the disease demands. It works. The catch is who it works for.

More than 90% of people who take insulin do not use a pump at all. They use a pen. It's discreet, it's familiar, it fits in a pocket, and it doesn't tether you to a device you wear every minute of every day. The pen population is the overwhelming majority - and the automation revolution largely passed them by. The industry's answer was usually some version of: convert them. Get them onto a pump.

Luna's founders found that answer a little absurd. If most people have already voted with their pockets for the pen, maybe the product should meet them there.

Three people who already built the things you use

This is not a team learning diabetes on the job. John Sjolund was diagnosed with type 1 at age 4 and went on to create Timesulin, the smart cap that told pen users when they last dosed. Sean Saint was an early Dexcom developer who founded Companion Medical and built InPen, the connected insulin pen that a large share of pen users already know. Jon Brilliant helped build Welldoc's BlueStar and co-founded Bigfoot Biomedical, later acquired by Abbott.

Between them, they have shipped some of the most widely used diabetes products in the world. So when they came together in 2020, the bet they made carried weight: that the biggest unclaimed prize in diabetes was not a better pump, but automation that pen users would actually accept. And they narrowed it further to a single, almost cheeky insight - you don't need to automate the whole day to capture most of the benefit. You need to automate the night.

Pen by day. Automation by night.

Luna is a hybrid by design. During the day, you keep your pen - the routine you already trust. At night, the small wearable patch pump takes over, working with a continuous glucose monitor and a fully closed-loop algorithm to deliver automated micro-doses of rapid-acting insulin while you sleep.

Why bother with only half the clock? Because the company points to a striking number: over 80% of the glucose improvement from traditional pump-based automated insulin delivery happens overnight. Sleep is when blood sugar drifts unsupervised for hours, when hypoglycemia is scariest, and when a person can do the least about it. Automate that window and you capture most of the upside - without asking anyone to wear a pump through every meeting, workout and shower.

A trial, a cap table, and a measurable promise

Skeptics of digital health have heard a thousand pitches about simplifying diabetes. Luna's answer is to point at the things that are hard to fake. Its pivotal trial uses Time in Range - the increasingly standard yardstick for glucose control - as the primary endpoint. Not engagement, not app downloads. The percentage of time a person's glucose actually stays where it should.

Then there's the money, and who's behind it. The $23.6M Series A was led by Vensana Capital, a dedicated medtech investor, with the Swiss Diabetes Venture Fund and Ascensia Diabetes Care - a global diabetes-care company - joining in. Winklevoss Capital, better known for crypto, is on the cap table too, which is either a sign of broad conviction or simply proof that good ideas attract odd company.

The partnership signals matter as much as the dollars. Ascensia's involvement plugs Luna into a company that already reaches people with diabetes worldwide - the kind of strategic backer you want when your plan involves manufacturing at scale and earning the trust of clinicians.

Automation for the masses, not the few

Luna states its mission plainly: to bring the transformative power of insulin automation to the masses. Read that against the backdrop of who currently benefits from automation, and it stops sounding like a slogan. The technology has existed for years. The gap has always been access - and access, in Luna's telling, is a design problem as much as a clinical one.

The company talks about human factors and user experience the way most medical device makers talk about specs. That emphasis traces straight back to the founders, who built their reputations not on inventing insulin or glucose sensing, but on making both less of a daily burden. The mission is less about a breakthrough molecule and more about a breakthrough in who gets to feel the benefit.

The most interesting hour is the one you sleep through

Diabetes does not clock out at night. It is a 24-hour disease managed by people who would very much like to spend some of those hours not managing it. Luna's wager is that the future of insulin automation isn't an all-or-nothing choice between a pen and a pump - it's a system that knows when to step in and when to stay out of the way.

There's real risk here. The product is investigational, the trials must read out, and regulators decide the timeline, not press releases. Luna is a small team taking on entrenched device makers with a deliberately narrow first act. But the narrowness is the point. By refusing to automate everything, it might finally automate the part that matters most - for the most people.

So return to that bedroom in San Diego. The patch is still working. The person is still asleep. Tomorrow they'll pick up their pen and go about their day, the same pen they always used, never having joined the small club of full-time pump users. The only thing that changed is the night - quietly, automatically, for once not their problem to solve. If Luna is right, that's not a small thing. For millions of people, it's the whole point.