Most small teams ship a product and then quietly make the person who built it the pager. We offer the other option: an MLabs engineer who owns keeping your product running, on your own infrastructure, for $600 a month.
The person who built it becomes the on-call rota
After launch the work changes. Someone has to notice the certificate that expires, the disk that fills, the scheduled job that stopped running on Saturday. On a team of three that person is usually whoever wrote the code, and they also have a roadmap. Alerts go to a channel nobody reads at the weekend, or to one phone that is never off.
That arrangement holds until the week it doesn't. Then the roadmap stops while the builder works out what broke.
Setup turns what you run into alerts worth acting on
We start with an inventory of what's live: services, databases, scheduled jobs, certificates, and on-chain state if your product has contracts. Then we write the alert definitions, stand the checks up on your infrastructure and tune them for two weeks, so the alerts that fire are the ones that need a person. You get a one-page runbook at the end.
Setup is a one-off $1,000. We host nothing. Everything runs in your accounts and stays there if you stop working with us.
Level 1: the checks run all day and an engineer fixes what they find
For $600 a month the checks run around the clock without anyone watching them. When one fires, it reaches us. An MLabs engineer responds the next business day, in CET hours, and has five hours a month for fixes and tuning. Each month you get an uptime report and an incident report. You can cancel at any month end.
At 3am nothing human happens until morning
If your API falls over at 3am on a Sunday, the automation sees it, records it and pages us. A person picks it up on Monday morning CET. Levels 1 to 3 all work this way. For a product whose users can wait until morning, that costs a fraction of having a person awake all night.
If your product can't wait until morning, that's Level 4.
Higher levels add a named engineer, more hours, then round-the-clock cover
Level 2 gives you a named engineer and a backup, with more hours for responding to alerts. Level 3 adds a monthly basket of engineering hours for features and updates, scheduled and reported each month. Level 4 puts a human on call 24 hours a day. We quote those after a conversation, because the price depends on what you run.
We run our own services the same way
MLabs runs its own production services under the same routine. That includes OpenBao, which has held the secrets for our own services since 21 October 2026.
If your product is live and your team is too small to staff it, tell us what's running. We'll ask about your stack and say which level fits.
