No-Code Semantic Layer for Finance
Preql
This is really a distribution advantage for Preql, because the hardest part of a semantic layer is not calculating metrics, it is getting the right person to define and maintain them. dbt put that job with analytics engineers inside a code workflow, and Cube started from a developer model too. Preql is aiming at finance teams whose real source of truth often still lives across spreadsheets, warehouse tables, and manual reporting steps, so a no code interface lowers the cost of getting those definitions into the system.
-
dbt’s semantic layer came from the Transform acquisition and is powered by MetricFlow, with metrics defined in YAML and SQL adjacent workflows. That fits teams already living in dbt, but it assumes someone technical owns business logic as code, which is natural for analytics engineering and less natural for finance operations.
-
Preql is built around the opposite starting point. Its product tries to pull logic out of offline finance workflows, especially spreadsheet based reconciliations and inconsistent KPI definitions, then turn that into a governed semantic model with agents and business owner approval loops. That maps to how finance teams actually work before a clean warehouse exists.
-
Cube has moved toward broader accessibility with Visual Model and AI model generation, but its center of gravity is still a semantic layer platform that grew up with developers and embedded analytics teams. Preql is more opinionated around finance as the buyer and around replacing manual reporting work before expanding into broader AI and BI use cases.
The category is moving toward tools that let business owners shape metric definitions directly, while still keeping technical guardrails underneath. If that shift continues, the winners will be the products that can turn messy spreadsheet era business logic into reusable system logic without requiring finance teams to first become dbt users or semantic model developers.