The Scenario
Every quarter you run a revenue reconciliation. The deliverable is simple: the Q4 Revenue dataset in Databox gets replaced with fresh, audited data pulled from the finance workbook. The board dashboard reflects the final numbers. Nothing carries over from Q3.
The process is not simple. The old dataset has to be deleted first — Databox doesn't support in-place replacement for custom datasets. Then a new one has to be created with the same primary key structure. Then 400 rows from the RevenueClean worksheet have to be pushed in, formatted exactly the way Databox expects, with deal_id as the primary key.
You've done this before. Last quarter the delete step alone took 15 minutes because you had to navigate three levels into Databox's Data Manager, confirm through two dialogs, and then manually re-enter the schema for the new dataset field by field. The push step required a CSV export, a round of cleanup to remove the Excel currency formatting that finance uses, and a second import attempt after the first one failed on 8 rows.
The bad version:
- Navigate to Databox Data Manager, find the 'FY2024 Revenue' dataset, confirm the delete through two confirmation dialogs.
- Manually re-enter the schema for the new dataset — field names, types, primary key — because Databox doesn't carry it forward.
- Export the RevenueClean sheet as a CSV, strip currency symbols and Excel thousand-separator formatting, and attempt the import.
- Get a partial success: 388 out of 400 rows loaded, with no clear error message explaining which 12 failed or why.
Quarter-end has enough moving parts without spending two hours on this.
The Easy Way: One Prompt in SheetXAI
SheetXAI is an AI agent that lives inside your Excel workbook. It reads the RevenueClean worksheet, deletes the old Databox dataset, recreates it with the correct schema, and pushes all 400 rows — including stripping any formatting that would fail the validator.
Open the SheetXAI sidebar and paste this:
In Excel, read the RevenueClean sheet, delete the old Databox dataset named 'FY2024 Revenue', recreate it with the same primary key (deal_id), and push all 400 rows — strip any currency symbols or Excel number formatting from the revenue column first.
SheetXAI confirms the delete, creates the new dataset with deal_id as the primary key, cleans the revenue column, and pushes all 400 rows as a batch. Row-level status is written back to column H of the RevenueClean worksheet.
What You Get
- Confirmation that the old 'FY2024 Revenue' dataset was deleted, including its dataset ID.
- A new dataset created with deal_id as the primary key and the same schema as the original.
- All 400 rows pushed, with a success count and any skipped rows logged with reasons in column H.
- Revenue column cleaned of currency symbols and Excel formatting artifacts before the push — no silent type mismatches.
What If the Data Is Not Quite Ready
The workbook has rows marked "draft" that shouldn't go to Databox
Preliminary deals that haven't closed. They're in the worksheet but not in the reconciliation.
Delete the Databox dataset named 'FY2024 Revenue', create a new dataset with primary key 'deal_id', filter the RevenueClean sheet to only rows where column F (status) equals 'closed', then push the filtered rows to the new dataset.
The revenue column uses mixed currency symbols from different regional teams
Some rows have "$" amounts, others are "€" with European decimal formatting.
In the RevenueClean sheet, normalize the revenue column (column D) by stripping currency symbols and converting European decimal formatting (period as thousands separator, comma as decimal) to standard US format, then delete Databox dataset 'FY2024 Revenue', recreate it with primary key 'deal_id', and push all 400 rows.
The Q4 data lives in two worksheets — one for enterprise deals, one for SMB
Both need to land in the same Databox dataset.
Join the EnterpriseQ4 sheet and SMBQ4 sheet on deal_id, write the merged result to a ReconciliationFinal sheet, then delete Databox dataset 'FY2024 Revenue', create a new dataset with primary key 'deal_id', and push all merged rows from ReconciliationFinal.
You want to archive the old dataset before deleting it
Finance policy requires a snapshot before any delete operation.
Create a new Databox dataset called 'FY2024 Revenue — Archived' under the same data source and push all current rows from the RevenueClean sheet to it as a backup, then delete the original 'FY2024 Revenue' dataset, recreate it with primary key 'deal_id', and push all 400 rows from the RevenueClean sheet.
One prompt handles the archive, the delete, the recreate, and the refresh — the audit trail exists and the dashboard is current before you close the laptop.
Try It
Get the 7-day free trial of SheetXAI and open your quarter-end finance workbook, then ask it to handle the delete-and-repopulate cycle against your Databox account. You can also see how to initialize a Databox data source from a workbook schema or view the full Databox integration overview.