Who hurts
Data teams at 50–500 person companies with more than one BI tool.
The problem
The same metric is defined four times across dashboards, and the definitions drift. Meetings become arguments about whose number is right, and the data team's credibility erodes.
What you build
Parse the SQL behind every dashboard, cluster near-identical metrics, surface the conflicts with both definitions side by side, and alert when a definition changes under a published dashboard.
Why now
SQL parsing plus semantic comparison finally makes 'these two metrics are almost the same' detectable rather than folklore.
Validate it this week
Run the parser over one company's BI export and show them their four definitions of revenue. The room goes quiet — that's the sale.
Why you'd keep winning
Parser coverage across BI tools and warehouse dialects. Tedious, cumulative, and hard to catch up on.
The honest risk
dbt and the semantic-layer vendors are moving into this space. Be the tool for companies who never adopted a semantic layer.
The prompt is written to make an AI argue with you before it writes code — that first round of pushback is worth more than the scaffold.