We spent 10 years in data teams. Here's what kept us up at night.

Every Monday morning, somewhere in a company, someone asks: "Why did revenue drop over the weekend?"

And the data team scrambles. They pull logs, query the warehouse, build a one-off chart. They find the answer: it was a payment gateway issue, three days ago. The business lost money for 72 hours before anyone noticed.

Then someone says: "We should set up monitoring for this." And they do. A dashboard gets built. An alert gets configured. Six months later, the same thing happens again, because the dashboard rotted, the alert was too noisy, or the engineer who set it up left.

We watched this loop play out at every company we worked at (fintech, gaming, e-commerce), it didn't matter. The problem wasn't that teams didn't care about their metrics. It was that reliable monitoring was always an engineering project. SQL to write. Thresholds to tune. Infrastructure to maintain. Most teams just chose not to do it, until something broke.

So we built Lighthouse. Describe a metric in plain English. Connect your warehouse in 10 minutes. Get a Slack alert the moment something moves. No SQL required, no engineering tickets, no dashboard that nobody looks at.

This is the tool we wish we'd had. We hope it's useful to you.