Chat
Answer questions with shared business logic.
Models
Transform raw data and define reusable measures and segments. Basedash AI builds the model, then every AI workflow can use it.
14-day trial. No credit card required.
Models / Customer activity
Ask AI
Models
Customer activity
models.customer_activity
Subscriptions
models.subscriptions
Qualified pipeline
models.qualified_pipeline
Product usage
models.product_usage
Customer activity
One row per customer event with account and plan context.
01 select e.user_id, e.occurred_at, e.event_name,
02 a.plan, a.region, a.created_at as account_created_at
03 from events e join accounts a on a.id = e.account_id
04 where not a.is_internal
reference
customer_activity
measures
3
segments
4
One layer
Shape raw data and define its business meaning in the same place.
Models turn raw warehouse tables into analysis-ready datasets. Each model combines reusable SQL with documented dimensions, aggregate measures, reusable segments, and a default time column, giving people and AI one complete description of how your business data should be queried.
Because Models are built into Basedash, every AI can use the same transformation and semantic context in chat, dashboards, insights, and automations. It is a first-party alternative to stitching together dbt for transformation and Cube for semantic modeling—without separate infrastructure or another deployment workflow.
Why it matters
Query the same model from every chart, dashboard, agent, and SQL workflow.
01Model once
02Reference anywhere
Subscriptions
v12select subscription_id, account_id,
started_at, status, plan,
monthly_amount
from subscriptions
where not is_test
Verifiedmodels.subscriptions
Net revenue dashboardRan 2m ago
select date_trunc('month', started_at) as month,
sum(monthly_amount) as mrr
from
models.subscriptions
where status = 'active'
group by month
$59K+24%($48K)
Coverage
Dimensions, measures, and segments travel with every Basedash workflow.
Charts
Generate visualizations from modeled data.
Dashboards
Keep every report on the same measures.
Insights
Spot trends using consistent segments.
Automations
Schedule reports from trusted models.
SQL editor
Query models like regular virtual tables.
Examples
Create clean datasets with the measures and segments every team needs.
Models
3
Subscriptions
RevenueOne row per subscription with account, plan, status, and billing details.
4 measures · 3 segments2d ago
Customer activity
GrowthClean customer events joined with account and plan attributes.
3 measures · 4 segments5d ago
Product usage
ProductFeature usage at the event grain with documented product dimensions.
5 measures · 2 segments1w ago
Governance
Transformation, semantics, governance, and AI creation in one product.
Models centralize the datasets and business logic your organization depends on. Teams can review usage, document every field, restore earlier versions, and verify approved models while existing data-source access controls continue to apply.
| Governance capability | External modeling stack | Basedash Models |
|---|---|---|
| Transformation | Build and deploy transformations in a separate tool. | Create reusable row-grain datasets directly in Basedash. |
| Semantic modeling | Configure measures and segments in another semantic system. | Define dimensions, measures, segments, and default time together. |
| AI creation | Write project files and expressions by hand. | Describe the model and AI creates the full configuration. |
| AI consumption | Integrate model metadata with each downstream AI. | Every Basedash AI receives the model catalog automatically. |
| Change history and audit | Coordinate versions across repositories and deployments. | Every SQL or semantic edit creates a restorable version. |
| Verification | Approval depends on external review workflows. | Admins verify approved models for people and AI. |
| Access scope | Recreate data access in the modeling layer. | Existing data-source and object access controls carry through. |
| Operations | Maintain more infrastructure and deployment pipelines. | Use native SQL with no separate modeling service to operate. |
Models use your data source’s native SQL dialect. Read the Models documentation to see how measures, segments, verification, and version history work.
Models, answered.
What are Basedash Models?
Basedash Models are reusable SQL datasets with semantic metadata. Each model belongs to one data source and can include documented dimensions, aggregate measures, reusable segments, a default time column, verification, and version history.
Can AI create a complete model?
Yes. Describe what you want to model, including business concepts such as active users or retention rate. Basedash AI can inspect your schema and create the model SQL, column descriptions, measures, segments, and default time column, then explain what it built.
Can Models handle data transformation?
Yes. A model can clean, join, rename, and reshape warehouse data into a reusable dataset. Row-grain output is usually the most flexible because measures and segments can then adapt it to many analyses, while fixed pre-aggregated output remains available when you intentionally need it.
What are measures and segments?
Measures are reusable aggregate SQL expressions, such as total revenue or active users. Segments are reusable predicates, such as active customers or completed orders. Basedash AI places measure expressions in SELECT and segment expressions in WHERE using normal SQL.
How do I query a model?
Query a model like a read-only virtual table using models.<reference_name>, such as FROM models.subscriptions. Join, filter, aggregate, and group it with the native SQL dialect of its data source.
How are Models different from skills?
Models contain executable SQL and structured semantic metadata. Skills are reusable prose instructions for AI behavior. Use Models for datasets, measures, and segments; use skills for broader analysis methods, preferences, and business guidance.
Can Models replace dbt or Cube?
Models provide first-party transformation and semantic modeling inside Basedash, so teams can create reusable datasets, measures, and segments without operating a separate dbt or Cube deployment. They use the data source's native SQL and are available to every Basedash AI workflow automatically.
Can Models be audited and verified?
Yes. SQL and semantic changes create restorable model versions, and the Used by view shows dependent models, charts, variables, and automations. Organization admins can verify approved models, and semantic changes remove verification until the new version is reviewed.