Use case

Bring mismatched spreadsheets into one usable table.

Channel exports differ in headers, dates and field definitions, with duplicate records mixed in. Define matching keys and rules, then consolidate data while retaining exceptions and sources.

A normalized table, exception list and links back to source records.

Start your work
Illustrative workflow with fictional data and original vector artwork; not a live customer task, generation result or external action.

Combine these order files into order ID, SKU, quantity, date and amount.

This page focuses on one task: A normalized table, exception list and links back to source records.

Start with these inputs.

Excel/CSV inputs, target fields, matching keys and deduplication rules.

Retain the originals

Keep source images, text or file references beside the processed results for comparison.

Set the requirements

Confirm ambiguous dates, identifiers and matching rules instead of silently merging distinct records.

Illustrative workflow with fictional data and original vector artwork; not a live customer task, generation result or external action.

Work through the task in three steps.

01Define field mappings and matching rules.

Excel/CSV inputs, target fields, matching keys and deduplication rules.

Example prompt
Combine these order files into order ID, SKU, quantity, date and amount. Check duplicates by order ID and SKU. List possible duplicates first and retain sources rather than deleting them immediately.
02Normalize formats and identify possible duplicates or gaps.

Channel exports differ in headers, dates and field definitions, with duplicate records mixed in. Define matching keys and rules, then consolidate data while retaining exceptions and sources.

03Confirm treatment and retain both clean data and an audit record.

Confirm ambiguous dates, identifiers and matching rules instead of silently merging distinct records.

Deliverable

A normalized table, exception list and links back to source records.

Keep the results and review evidence together.

A normalized table, exception list and links back to source records.

Revise the part that needs work

Identify the row, cell or result to change and review that part while retaining the rest.

Review before use

Confirm ambiguous dates, identifiers and matching rules instead of silently merging distinct records.

Illustrative workflow with fictional data and original vector artwork; not a live customer task, generation result or external action.

Before you begin

What should I provide?

Excel/CSV inputs, target fields, matching keys and deduplication rules.

What needs review?

Confirm ambiguous dates, identifiers and matching rules instead of silently merging distinct records.

Is the demonstration a live task result?

Illustrative workflow with fictional data and original vector artwork; not a live customer task, generation result or external action.

Does viewing this page run a task or incur a charge?

This demonstration does not execute a task. Work starts after you submit a request in the workspace, subject to the actual configuration, approval requirements and costs.

Bring this task to the Agent.

Excel/CSV inputs, target fields, matching keys and deduplication rules.

Open TableRun