Projects / 2026
Running a beekeeping SaaS end to end, alone
Fifty to sixty two percent. That's how many colonies American beekeepers lost in the 2024 to 2025 season, depending on whether they run it as a hobby or a business. The worst numbers anyone has on record, and most of it traces back to something that was visible weeks earlier in a mite count nobody plotted.
That's the whole reason Beekio exists.
The loop is deliberately small. You add your hives, you log an inspection, and you get back a prioritized list of what to do in the next 24 to 48 hours for each one. Varroa count, brood pattern, weight, a few notes. The logging has to be fast enough to do one handed while wearing gloves, standing over an open hive, and that turned out to be the constraint that shaped most of the interface. On the Pro and Scaling plans the plan generates itself at 5 a.m. local time, so it's already waiting when you head out. There's a chat too, Ask Beekio, that knows your own hive history instead of answering from general knowledge. Every inspection gets stamped with the weather where the hive actually sits.
Most beekeeping apps are logbooks. They help you write down what happened. Beekio is trying to tell you what to do next, and that difference is the entire bet.
It's an expensive bet. Pro is 36 dollars a month in a market where HiveTracks charges 6.99 and HiveBook gives you unlimited hives for nothing. I've read enough forum threads to know exactly how that lands. The honest answer isn't that the competition is bad, it's that one hive you don't lose pays for a year of Pro several times over, and I have to prove that rather than assert it. Nobody in beekeeping owes me the benefit of the doubt.
Under the hood it's a normal indie SaaS stack, and I mean that as a compliment. A FastAPI backend on Railway, Neon for Postgres, Clerk for auth, Stripe for billing, Resend for transactional mail, and Anthropic's models doing the actual reasoning. Nothing exotic. Every hour I don't spend building my own auth is an hour I spend on the part nobody else can build for me.
The breakages are the part worth writing down.
The Clerk webhook spent longer than I want to admit pointing at the wrong URL. That meant people signed up fine and then didn't exist in my database, a failure that looks like nothing at all until you go looking for a row that should be there.
Invitation emails went to spam for days. Not a code bug in the slightest. DNS records, sender reputation, the whole boring unglamorous layer, fixed eventually by moving mail for beekio.com over to Proton and grinding through every record until it verified.
Then there was a script called send_test_nurture.py, written to test the waitlist sequence. It wasn't in git, it was sitting on a detached HEAD on exactly one machine, and on first run it fired a real production email through Resend with no confirmation prompt. It worked precisely as designed, which was the problem. That one goes in the same drawer as everything else I've built that was one keystroke away from being embarrassing.
Right now Beekio is pre launch. The site is a waitlist, not a checkout. Every plan tier ends in Join the Waitlist rather than Buy, and I'd rather say that plainly here than have you click through expecting something you can pay for today. The custom sign in pages are still on the dev branch because main is protected and I haven't merged them. A beta trial tier is waiting on a SQL migration I want to read twice before I run it.
The contact page says Beekio is a small team, and that every message is read by someone who works on the product, usually the same person who built it. That's a nice sentence. It's also just true, and it's why the roadmap is sequenced instead of parallel. One person can only break one thing at a time.
If you keep bees, or you know someone who does, it's at beekio.com. The waitlist is open.