AI-ASSISTED PIPELINE DEVELOPMENT
Build pipelines with AI, code, and SQL.
Describe a workflow in DDL’s interface. Use AI to generate pipeline logic in a high-level language, then write SQL directly when the task calls for query-level control.
UI-led workflowsAI-generated logicDirect SQL queries
AI pipeline builderExample workflow
DESCRIBE YOUR GOALPrepare a weekly delivery review by region.
SQL remains availableWrite direct queries and transformations in the same data workflow.
FROM GOAL TO PIPELINE
Start with the task. Review the logic.
Move from a plain-language workflow goal to a proposed pipeline, then inspect the steps before deciding how to proceed.
- 01Describe the workflow
Explain the data task in the DDL interface and provide the context it needs.
- 02Generate pipeline logic
Use AI to create high-level pipeline logic from the requested outcome.
- 03Review and refine
Inspect the proposed steps, make adjustments, and decide what to use.
AI-generated logic is a starting point. Review the proposed workflow before using it.
DIRECT SQL
Use SQL when you want query-level control.
Pipeline generation and SQL work side by side: start from a broader workflow with AI, or write a query directly for focused filtering, joining, and aggregation.
SQL QUERY WORKFLOW
Keep the query close to the work it supports.
Write SQL in the DDL workspace, inspect the result, and carry the query into the wider pipeline when it fits.
The statement shown is an example and has not been executed. Table names and behavior depend on the connected source and DDL configuration.
delivery_review.sqlExample SQL · not executed
SELECT
region,
COUNT(*) AS delayed_orders
FROM sample_delivery_events
WHERE delivery_status = 'Delayed'
GROUP BY region
ORDER BY delayed_orders DESC;
region,
COUNT(*) AS delayed_orders
FROM sample_delivery_events
WHERE delivery_status = 'Delayed'
GROUP BY region
ORDER BY delayed_orders DESC;
Example outputRun in a configured DDL project
AI-CENTERED, HUMAN-REVIEWED
Keep the work connected to business context.Delman has worked with enterprise data since 2018. DDL is available on-premise; model routes, permissions, and data paths depend on each organization’s configuration.
AI AT THE CENTER