Daniel Reinhardt

I got my pre-diabetes diagnosis two weeks after my 34th birthday, sitting in a Prenzlauer Berg doctor's office that still used paper charts. My HbA1c was 6.3%. The doctor suggested "maybe less bread" and sent me on my way. I'm a machine learning engineer—I instrument distributed systems for a living, tune latency down to microseconds, obsess over measurement error in training pipelines. The idea that I was supposed to fix my metabolism with vibes and portion control felt insane. So I did what I do professionally: I started collecting data. Freestyle Libre, then Dexcom. Oura ring. Whoop band for a few months until the subscription model annoyed me. Quarterly bloodwork panels that my Techniker Krankenkasse definitely does not think are necessary. I logged everything in spreadsheets, ran correlations, learned that my glucose spikes hardest at the kombuchas I thought were healthy and barely moves after full-fat Greek yogurt. N=1, obviously. Your pancreas is not my pancreas.

What surprised me wasn't the data—it was how bad most of these products are. The apps crash. The sensors fall off in the shower or just lie to you for the first 24 hours. Customer support is either non-existent or handled by someone reading from a script in Manila who can't escalate anything. I started writing reviews because I was already in the Notion docs anyway, and a friend suggested I put them somewhere public. I'm not trying to build a brand. I don't do affiliate links. I buy everything myself, and if a company sends me something unsolicited, I'll say so in the review and probably roast them for the cold email.

My wife is seven months pregnant now, and I've realized this quantified-self thing is about to get significantly more complicated. I'll probably end up reviewing baby thermometers with the same energy I brought to continuous glucose monitors, which is either good content or a sign I need therapy. Maybe both. I work from our altbau in Friedrichshain, usually standing, usually with cold coffee I forgot about three hours ago. If you're here for influencer enthusiasm or clean affiliate income, you're in the wrong place. If you want someone to actually test whether that $400 wearable does what it claims—with screenshots, error bars, and footnotes—then yeah, stick around.