The Basedash semantic layer lets teams create reusable SQL metrics and models that AI can reference across chat, charts, dashboards, insights, and automations.
Hey everyone, Max here from Basedash.
Today we're launching the Basedash semantic layer: reusable SQL definitions for AI analytics in Basedash.
These definitions are reusable SQL queries attached to a data source. You can define "monthly recurring revenue", "activation rate", or "qualified pipeline" once, give it a reference name and description, and Basedash can use that exact SQL anywhere.
This matters because AI is great at exploring data, but teams still need deterministic calculations for the metrics they run the business on. With definitions, the AI can build charts, answer chat questions, generate insights, and run automations while reusing the same approved SQL every time.
We've been using this internally for metrics that show up across dashboards and reporting workflows. It removes the copy-paste SQL problem and makes the AI's work easier to audit.
The semantic layer is available in Basedash today. PH community gets an extra week on the trial this week. Happy to answer anything.
Report
@maxmusing Congrats on the launch. Just a short question: when the AI uses your sematic layer definitions, how does it handle edge cases where the same metric might need slight variations across different contexts, say currency conversions, time zones or business units?
@swati_paliwal thanks! The metrics in your semantic layer act as a base, and you can easily tweak them for your specific needs when building reports. Just ask the AI “convert this to CAD for our Canadian client” and the AI will handle the variation for you.
Report
This feels practical for BI. AI can help people ask questions faster but the important business metrics still need one agreed definition. I like the idea of defining smth like MRR or activation rate once, then letting AI reuse that same logic everywhere. How do you handle cases where diff teams define the same metric slightly differently?
@ada_johnsen great question. Each team can (optionally) have their own set of metrics, and you can even define a metric using another team’s metric as a base. We support team-level AI context to teach the AI exactly how your team works.
Report
Semantic layers seem straightforward until different teams start defining the same metric differently.
Have you found the technical challenge is the easy part, and the harder problem is getting organizations to agree on shared definitions in the first place?
People love what AI does with their data right up until the AI gets something wrong. Which happens all too often with other tools, unfortunately.
That's why we're so focused around better context for AI agents, and why we think we're building one of the most accurate data agent platforms in the world today. Definitions take that even further. The metric gets written once, reviewed, and the AI reuses that exact SQL every time it touches it. So when finance and the AI both report revenue, it's the same revenue.
Replies
Basedash
@maxmusing Congrats on the launch. Just a short question: when the AI uses your sematic layer definitions, how does it handle edge cases where the same metric might need slight variations across different contexts, say currency conversions, time zones or business units?
Basedash
@swati_paliwal thanks! The metrics in your semantic layer act as a base, and you can easily tweak them for your specific needs when building reports. Just ask the AI “convert this to CAD for our Canadian client” and the AI will handle the variation for you.
This feels practical for BI. AI can help people ask questions faster but the important business metrics still need one agreed definition. I like the idea of defining smth like MRR or activation rate once, then letting AI reuse that same logic everywhere. How do you handle cases where diff teams define the same metric slightly differently?
Basedash
@ada_johnsen great question. Each team can (optionally) have their own set of metrics, and you can even define a metric using another team’s metric as a base. We support team-level AI context to teach the AI exactly how your team works.
Semantic layers seem straightforward until different teams start defining the same metric differently.
Have you found the technical challenge is the easy part, and the harder problem is getting organizations to agree on shared definitions in the first place?
Basedash
@samyak_sanklecha we’ve found that most great companies have centralized teams responsible for defining these kinds of metrics for the other teams.
Mailwarm
Congratulations on your launch. I like the idea.
Basedash
Thanks @thamibenjelloun!
Basedash
Wanted to add the why behind this one!
People love what AI does with their data right up until the AI gets something wrong. Which happens all too often with other tools, unfortunately.
That's why we're so focused around better context for AI agents, and why we think we're building one of the most accurate data agent platforms in the world today. Definitions take that even further. The metric gets written once, reviewed, and the AI reuses that exact SQL every time it touches it. So when finance and the AI both report revenue, it's the same revenue.
Give it a spin and tell us where it breaks :D
Basedash
@kris_lachance context is all you need