Notes on databases, AI, and building QueryBear.
I've been connecting my coding agents to everything: Datadog logs, Linear, Slack. It still feels like I hit the same wall every time I actually need to debug something serious. The model can read the...
I've been thinking lately about a line people don't love to cross: giving more people access to the database. There's still a stigma, like it's a holdover from when "database access" meant shared admi...
Every company says they're "data-driven." Very few actually are. The reason is simple: data is locked behind technical skills that most people in the company don't have. Think about how information fl...
There's a spectrum between "AI writes everything" and "humans write everything." The sweet spot for SQL is somewhere in the middle, and I think most people misunderstand where. Fully automated SQL gen...
Data teams at growing companies are drowning, and it's not because they're bad at their jobs. The structure is broken. Here's the typical setup: a company has 50-200 employees and a data team of 2-4 p...
I'm building an AI SQL tool, and I'm going to say something that might sound counterintuitive: you should still learn SQL. AI is excellent at translating your intent into a query. But you need enough...
I get this question a lot. The short answer: not entirely, but for many teams, yes. Metabase is great software. It's open source, it has a solid query builder, and dashboards work well. But I've watch...
For QueryBear's AI to write accurate SQL, it needs to know your database schema — table names, column names, types, and relationships. Here's how that works under the hood. Initial sync. When you conn...
Every company builds internal tools. Admin panels, reporting dashboards, data export scripts. They work great at first. Then they break, and nobody fixes them. I've seen this cycle at every company I'...
When I first built QueryBear, the core feature was AI-generated SQL. But the feature that actually changed how I work day-to-day is much simpler: saved queries. Here's what my workflow used to look li...
Every company says they're "data-driven." Most aren't, and the reason is simpler than you'd think: people can't access the data. The typical setup looks like this. Data lives in a Postgres or MySQL da...
I've been writing SQL since college. For years I worked at companies where the same pattern repeated: someone on the marketing or ops team needed data, they'd Slack an engineer, and that engineer woul...
