Skip to Content
AgentBuilding Transformations
How-to guideAgentAvailable

Building Transformations

Use the Agent to design, validate, and save transformations while keeping inputs, outputs, assumptions, and failure behavior explicit.

Last reviewed July 13, 2026

The Agent can help create supported transformations from a business requirement. Use it to inspect inputs, draft logic, preview results, and save a durable transformation—not to skip validation.

Before you start

  • Identify the business outcome and expected output grain.
  • Choose the source tables and stable join keys.
  • Decide whether SQL, Python, a notebook, grouping rules, or another supported transformation type is appropriate.
  • Define the output table or reusable property and how downstream consumers will use it.

Build a transformation

  1. Ask the Agent to inspect candidate inputs and schemas.
  2. Confirm the selected fields, grain, timezone, and identity keys.
  3. Describe the calculation and edge cases in business terms.
  4. Ask for a preview on a small, representative sample.
  5. Compare the preview with hand-checked examples.
  6. Review the generated code, configuration, and output schema.
  7. Save the transformation only after the preview and assumptions are correct.
  8. Run it and confirm the materialized output in Activity Logs and the warehouse.

Prompt example

Build a customer-level SQL transformation with one row per customer. Calculate completed-order revenue, refund value, order count, first order date, and most recent order date in the workspace timezone. Exclude test orders and show five hand-checkable customers before saving.

Validate the output

CheckQuestion
GrainIs there exactly one row per intended entity?
KeysAre IDs stable, non-null, and unique where required?
TimeAre timezone and date-window boundaries explicit?
MoneyAre currency, tax, discounts, and refunds treated consistently?
NullsDoes missing data produce the intended fallback?
RerunsIs replace, append, or merge behavior safe?
ConsumersWill reports, audiences, or destinations read the expected schema?

Edit and rerun

When a preview fails, ask the Agent to explain the first root error before changing code. Fix one cause at a time, rerun the smallest useful sample, and recheck the output contract after every schema change.

Need help?

Include your workspace, integration or job ID, and the first error message when you contact support.

support@vendodata.com
Last updated on