Kimi Sheets: Create, Download & Verify Spreadsheets

Product and workflow review: October 12, 2026. Privacy and workspace guidance checked: October 15, 2026.

How to create and verify a spreadsheet with Kimi Sheets: describe the spreadsheet you need, upload source files if useful, let Kimi Sheets build the workbook, preview the structure, download the XLSX file, then verify formulas, data, charts, pivot tables, assumptions, and source links before using it for real work.

Kimi Sheets can help you create structured Excel files from natural-language instructions, uploaded spreadsheets, CSV files, PDFs, or other source material. But verification is still your responsibility. Treat Kimi Sheets as an AI spreadsheet agent that can build the first version faster, not as an auditor, accountant, analyst, or final authority.

Quick workflow: create first, verify before trusting

The safest Kimi Sheets workflow has two phases. First, create the spreadsheet. Second, verify it before you rely on it.

Stage	What you do	Why it matters
1. Define the spreadsheet goal	Explain the workbook purpose, audience, sheets, columns, formulas, charts, and output format.	A precise request reduces invented structure and unclear assumptions.
2. Approve and upload source files	Confirm permission for the selected service and workspace, then add approved Excel, CSV, PDF, Word, images, or other supported source files.	Approval comes before disclosure; source files also reduce guessing and make verification easier.
3. Ask for a planned structure	Ask Kimi to show sheet names, columns, formulas, assumptions, and source fields before generating the final workbook.	You can correct the design before the file is built.
4. Generate and preview	Let Kimi Sheets create the workbook and preview the structure online.	Preview catches obvious layout, missing-sheet, or wrong-column problems early.
5. Download XLSX	Download the file and open it in Excel, WPS, or compatible spreadsheet software.	The local file is where you can inspect formulas, links, formats, and charts.
6. Verify manually	Check source data, formulas, totals, pivots, charts, formatting, assumptions, and outliers.	AI-generated spreadsheets can still contain wrong formulas, stale data, or unsupported assumptions.

What Kimi Sheets can do

Kimi Sheets is Kimi’s spreadsheet agent for Excel-style tasks. Kimi’s official pages describe it as a tool that can create structured Excel files, generate formulas, organize data into tables, build pivot tables, create charts, clean data, convert between spreadsheet and document formats, and work across multiple linked sheets.

Capability	What it means in practice	What to verify
XLSX output	Kimi Sheets can output spreadsheet files that you can download and continue editing locally.	Open the file, check tabs, inspect formulas, and confirm the file is not only a static table.
Natural-language spreadsheet creation	You can describe a workbook in plain English instead of writing every formula manually.	Check whether the generated structure matches your intended workflow.
Excel/CSV upload	You can upload an existing Excel or CSV file and ask Kimi to clean, merge, analyze, or restructure it.	Compare row counts, columns, totals, and sample records against the original file.
Formula generation	Kimi can create formulas such as lookups, calculations, totals, percentages, and linked-sheet formulas.	Inspect formula ranges and recalculate sample rows manually.
Multi-sheet workbooks	Kimi can create workbooks where sheets connect to each other through formulas and references.	Trace cross-sheet links and confirm formulas update when assumptions change.
Pivot tables and charts	Kimi’s product page describes pivot-table and chart generation for analysis and visualization.	Check pivot source ranges, aggregations, chart ranges, axes, labels, and filters.
Data cleaning	Kimi can help remove duplicates, normalize messy formatting, and organize raw data.	Run duplicate checks, missing-value checks, and category consistency checks.
Financial models	Kimi’s use-case library includes DCF valuation and financial modeling prompts.	Verify assumptions, formulas, discount rates, growth rates, source data, and model logic with a qualified person.

Before you start: prepare the spreadsheet brief

A better spreadsheet starts with a better brief. Do not begin with “make me a spreadsheet” unless the task is very simple. Tell Kimi what the spreadsheet is for, who will use it, what data it should include, and which parts must be calculated instead of hardcoded.

Goal: Define the business or personal decision the spreadsheet should support.
Permission and workspace: Confirm that your organization permits this data in the selected service. If Business controls are required, verify the active Enterprise Workspace before adding work data.
Source data: Use approved, clean Excel/CSV files where possible. If using PDFs or web sources, state what should be extracted and what should be ignored.
Workbook structure: Name the sheets you expect, such as Raw Data, Cleaned Data, Assumptions, Calculations, Summary, Dashboard, and Verification.
Columns: Specify required columns, data types, and formatting rules.
Formulas: Say which values should be calculated with formulas instead of typed as static numbers.
Charts and pivots: State which charts, summaries, filters, and pivot-table outputs you need.
Assumptions: Ask Kimi to separate assumptions from calculations and clearly label uncertain data.
Sensitive data: Remove unnecessary personal, client, financial, HR, medical, or confidential fields before upload.

How to create a spreadsheet with Kimi Sheets

Use this workflow when you want a real spreadsheet output rather than a text explanation of Excel formulas.

Work files and mobile access: The current Kimi Business platform guidance says the mobile app supports personal accounts, not Kimi Business. For data requiring Business controls, use a supported web or desktop session. Do not upload until the correct enterprise workspace and permission are confirmed; see the privacy checks below.

Open the Kimi service approved for your task and sign in.
Check the active workspace. For Kimi Business, select your organization’s Enterprise Workspace. On desktop, log in with the account that joined the organization, switch to Chat, then use the avatar/workspace menu to choose the enterprise workspace, as described in the official Business FAQ.
Click the Agent button at the bottom of the dialog box or choose the Sheets shortcut in the sidebar if it is available in your interface.
Enter a clear spreadsheet task in natural language.
Upload only approved source files if the workbook depends on existing data. Kimi’s general file support includes formats such as PDF, Word, Excel, PPT, images, TXT, and video, with stated limits of 100 MB per file and up to 50 uploaded files at a time.
Ask Kimi to show the planned workbook structure before final generation.
Review the planned sheet names, columns, formulas, assumptions, and source fields.
Ask Kimi to generate the workbook after you approve the structure.
Preview the spreadsheet online and look for obvious missing sheets, wrong columns, or unclear calculations.
Download the .xlsx file.
Open it locally in Excel, WPS, or another compatible spreadsheet editor and verify it before sharing or using it.

Master prompt for Kimi Sheets

This prompt is designed to make Kimi build a workbook that is easier to verify. Use synthetic data or files approved for your workspace; a prompt does not change account privacy settings.

Create an Excel workbook for this task:

Goal:
Create a spreadsheet that helps me [describe the decision or report].

Source data:
Use the uploaded files and only use web research if needed. If you use web sources, add source URLs and retrieval dates.

Workbook structure:
Create separate sheets for:
1. Raw Data
2. Cleaned Data
3. Assumptions
4. Calculations
5. Summary
6. Dashboard
7. Verification Notes

Rules:
- Use formulas instead of hardcoded totals where possible.
- Keep assumptions separate from source data.
- Highlight cells that need human review.
- Do not invent missing data. Mark missing or uncertain values clearly.
- Add notes explaining important formulas.
- If you create charts, state the source range and chart purpose.
- If you create pivot tables, state the source range and aggregation logic.
- Add a Verification Notes sheet listing what I should check manually before trusting the file.

Before generating the final XLSX, show me the planned sheet names, columns, formulas, assumptions, and source fields.

For important work, do not skip the “planned structure” step. It is easier to fix the spreadsheet design before Kimi generates the final workbook than after you have already downloaded a file full of wrong assumptions.

How to preview and download the XLSX

Kimi’s Help Center describes a workflow where you preview the spreadsheet structure online, then download the .xlsx file after confirming it is correct. During preview, check the workbook like a reviewer, not like a passive user.

Check that all expected sheets are present.
Check that columns are named clearly and consistently.
Check that raw data is not mixed with assumptions and calculations.
Check that summary tables match the spreadsheet goal.
Check that formulas are used where formulas are needed.
Check that “unknown” or missing values are marked instead of guessed.
Check that any web-sourced data includes source URLs or source notes.

After downloading, open the file locally. Some spreadsheet problems only become visible in Excel or another spreadsheet editor, especially formula references, chart ranges, pivot-table behavior, formatting, and cross-sheet links.

How to verify a spreadsheet created by Kimi Sheets

Verification is the most important part of the workflow. An AI-generated spreadsheet should not be trusted until you have checked its structure, data, formulas, summaries, and charts.

Check	How to verify it	What can go wrong
File opens correctly	Open the XLSX in Excel, WPS, or compatible software and save a clean copy.	Download may be incomplete, formatting may shift, or formulas may not behave as expected.
Sheet structure	Compare sheet names to your requested structure.	Missing sheets, extra sheets, or mixed raw/calculated data.
Row counts	Compare original file row counts with cleaned and summary sheets.	Rows may be dropped, duplicated, or filtered unintentionally.
Duplicate rows	Use spreadsheet duplicate checks on key fields such as ID, email, invoice number, SKU, or company name.	Duplicate leads, double-counted revenue, or repeated transactions.
Missing values	Filter blanks in required columns.	Blank values may distort totals, averages, charts, or pivots.
Date formats	Check day/month order, time zones, fiscal periods, and month labels.	Wrong date parsing can break trends and period comparisons.
Currency and number formats	Check currency symbols, decimal separators, percentages, and thousands separators.	Numbers can be treated as text, or percentages can be off by 100x.
Formula ranges	Click formulas and inspect referenced cells and ranges.	Formulas may skip rows, include headers, or reference the wrong sheet.
Manual recalculation	Pick 5–10 sample rows and calculate them manually with a calculator.	Formula logic may look correct but produce wrong results.
Cross-sheet links	Trace formulas that reference other sheets.	Linked assumptions may point to the wrong cell or stale value.
Pivot tables	Check source range, aggregation type, filters, and grouping.	Pivot tables may summarize the wrong range or use the wrong aggregation.
Charts	Check chart source ranges, axes, labels, legends, and filters.	Charts may show the wrong period, wrong category, or misleading scale.
Web sources	Open source links, confirm dates, and compare key figures to original pages.	Web data may be outdated, misread, or taken from a weak source.
Assumptions	Review the Assumptions sheet and check whether each assumption is reasonable.	AI may fill gaps with plausible but unsupported assumptions.

Formula verification: a practical method

Do not only look at totals. Open the formula bar and verify how totals were calculated.

Choose a calculated column.
Click the first calculated cell and inspect the formula.
Check whether the formula references the correct row and sheet.
Copy the logic into plain English: “Revenue equals units multiplied by price.”
Manually recalculate several rows.
Check whether the formula was filled down consistently.
Check whether blank rows, headers, or totals were accidentally included.
Change one assumption value and confirm dependent calculations update.

If the spreadsheet contains financial projections, DCF valuation, budgets, tax calculations, contract terms, compliance tracking, medical data, or academic results, ask a qualified person to review it before using it for decisions.

How to verify charts and pivot tables

Charts and pivots are useful only when they summarize the right data. A clean-looking dashboard can still be wrong.

Element	Verification step	Red flag
Pivot table source	Open pivot settings and confirm the source range includes all current rows and columns.	The pivot excludes new rows or includes blank rows.
Pivot aggregation	Check whether values are summed, counted, averaged, or grouped correctly.	Revenue is counted instead of summed, or rates are averaged incorrectly.
Pivot filters	Check date, category, status, and region filters.	A filter hides important rows without warning.
Chart source range	Click the chart and inspect selected data ranges.	The chart uses a partial range or old summary table.
Chart axis	Check axis start, scale, units, and labels.	A misleading scale exaggerates or hides change.
Chart type	Confirm the chart type fits the data.	Pie charts with too many categories or line charts for unordered categories.
Dashboard filters	Test filters and slicers if present.	Filters change one chart but not the matching summary table.

Prompt examples for Kimi Sheets

Use these examples as starting points with synthetic data or approved source files. For important spreadsheets, add your own data definitions and verification rules.

Use case	Prompt example
Monthly sales report	`Create a monthly sales report workbook from the uploaded CSV. Include Raw Data, Cleaned Data, Sales Summary, Regional Performance, Product Performance, Dashboard, and Verification Notes. Use formulas for totals and growth rates. Add charts for monthly trend and regional split. Do not invent missing sales data.`
Budget tracker	`Create a personal monthly budget spreadsheet with income, fixed expenses, variable expenses, savings, debt payments, category totals, and monthly surplus. Use formulas, conditional formatting for overspending, and a Verification sheet explaining what I should check.`
Project tracker	`Build a project tracker with task ID, task name, owner, priority, status, start date, due date, completion percentage, dependencies, and a Gantt-style timeline. Add a dashboard showing overdue tasks, status count, and workload by owner.`
DCF model	`Create a DCF valuation workbook for the uploaded company data. Separate assumptions, historical data, forecast calculations, valuation output, and verification notes. Use formulas for revenue growth, margins, free cash flow, discounting, terminal value, and sensitivity tables. Mark all assumptions clearly and cite any external data sources.`
Literature matrix	`Turn these research papers into an Excel literature matrix. Include title, authors, year, research question, method, dataset, key finding, limitation, and relevance. Do not summarize claims without noting which paper they came from.`
Lead list from PDF	`Extract company leads from this trade-show PDF into a clean XLSX file. Include company name, contact person, email, phone, website, product category, country, source page, and confidence note. Mark unreadable fields instead of guessing.`
Campaign performance report	`Create a campaign performance workbook from the uploaded ad data. Include spend, impressions, clicks, CTR, CPC, conversions, CPA, revenue, ROAS, and notes. Build a dashboard and flag campaigns with missing or unrealistic values.`
Merge monthly files	`Merge these 12 monthly Excel files into one annual workbook. Keep a Raw Data sheet for each month, create a consolidated table, calculate month-over-month growth, and add a Verification sheet listing row counts from each source file.`
Audit an existing spreadsheet	`Audit this spreadsheet and find likely formula, data-quality, duplicate, formatting, and chart-range issues. Do not modify the file yet. First give me a checklist of problems and the cells or sheets involved.`

Common errors and how to catch them

Problem	How it usually appears	How to catch it
Static totals instead of formulas	Totals look correct once but do not update when values change.	Click total cells and check whether they contain formulas.
Wrong formula range	One month, region, product, or row group is missing from totals.	Inspect formula references and compare row counts.
Duplicate records	Revenue, leads, or tasks appear inflated.	Run duplicate checks on key identifiers.
Missing values	Averages, totals, or pivot tables look unusually low.	Filter blanks in required fields.
Numbers stored as text	Formulas ignore values or sorting behaves oddly.	Check cell format and convert text numbers to numeric values.
Wrong date parsing	Monthly trends or timelines look incorrect.	Check date format, day/month order, and fiscal period mapping.
Assumptions hidden in formulas	Growth rates, discounts, or conversion rates are buried inside formulas.	Move assumptions into a separate sheet and reference them from formulas.
Misleading chart	A chart looks persuasive but uses the wrong range or scale.	Click the chart and inspect source range, axis scale, labels, and filters.
Outdated web data	External numbers do not match current sources.	Open source URLs and verify dates manually.

Kimi Sheets vs normal Kimi chat, Kimi Docs, and Kimi Slides

Choose the Kimi mode based on the output you need.

Tool	Best for	Not ideal for
Kimi Sheets	XLSX files, formulas, tables, dashboards, financial models, data cleaning, pivot tables, charts, and multi-sheet workbooks.	Long narrative documents or presentation decks when spreadsheet output is not needed.
Normal Kimi chat	Quick explanations, formula suggestions, and simple spreadsheet advice.	Building and downloading a full workbook that needs formula and sheet structure.
Kimi Docs	Word/PDF reports, document review, translation, contracts, written summaries, and formatted documents.	Spreadsheet calculations and dashboards.
Kimi Slides	Presentation decks, visual storytelling, and slide-based reports.	Data cleaning, pivot tables, and workbook formulas.
Kimi Deep Research	Source-backed research that may later feed a spreadsheet.	Final spreadsheet verification without manual checks.

Privacy, credits, and limits

Kimi’s general file-processing FAQ says it supports common file formats such as PDF, Word, Excel, PPT, images, TXT, and video, with each file no larger than 100 MB and up to 50 files uploaded at a time. Always check the live interface because limits and supported workflows can change.

Kimi’s membership documentation says features such as Agent, Deep Research, Slides, Docs, Sheets, Kimi Code, Kimi Work, and Kimi Claw share a single credit pool, and credits are consumed based on actual usage. Longer inputs, longer outputs, model choice, context length, and tool calls can affect consumption. Monthly and annual subscribers receive credits on a monthly cycle, and unused credits expire at the end of each cycle rather than rolling over.

Personal accounts and the applicable privacy policy

For accounts governed by the kimi.ai Privacy Policy, uploaded files, prompts, and generated content are covered as user content. The policy describes using that content to operate and improve the service, including model training. Product-specific policies can apply. Read the policy linked from your account rather than treating every Kimi service as one privacy environment.

The support process described in the kimi.com data-usage guidance is more specific: the page describes training after secure processing and de-identification, and account-level opt-out registration within 5–7 working days after identity verification. It does not offer a separate setting for one spreadsheet or conversation. Sending a request is not confirmation that opt-out has taken effect. Obtain confirmation before relying on it. Do not assume this procedure or timeframe applies to an account governed by a different policy; confirm the process with that service’s support.

Kimi Business: verify the Enterprise Workspace

Kimi Business documentation states that enterprise content is excluded from training by default and distinguishes the Personal Workspace from the Enterprise Workspace. Their data and benefits are separate. A personal paid plan, a work email address, or a professional task does not by itself establish Business coverage. Check that the Sheets session is in your organization’s workspace before uploading.

Business is not blanket permission to upload every regulated or confidential file. Ask your administrator to confirm the approved service, workspace, agreement, retention requirements, and any connected tools. If you cannot confirm the workspace or permission, stop before uploading or use synthetic data.

No training is not the same as no storage

Neither training opt-out nor the Business no-training commitment automatically guarantees zero retention or immediate deletion. The retention section of the kimi.ai policy allows different periods according to the data, service purposes, and legal or contractual requirements; deletion can have exceptions. Check the applicable agreement and any product-specific policy rather than treating a training control as a deletion setting.

Before uploading a sensitive spreadsheet:

Confirm permission for the exact data and service, and verify the Enterprise Workspace when Business controls are required.
Remove unnecessary names, IDs, addresses, signatures, account numbers, bank details, private keys, and client fields before upload. Review hidden sheets, comments, and metadata as well as visible cells.
Do not upload payroll, medical, legal, compliance, or confidential financial spreadsheets unless you have permission.
Ask for “missing” or “unknown” markers instead of letting AI guess sensitive or incomplete fields.
Keep an original copy of every uploaded source file and every downloaded workbook.
For business-critical spreadsheets, get human review before sharing, publishing, filing, or making decisions.

Troubleshooting Kimi Sheets outputs

Issue	Likely cause	Fix
Formulas are wrong	The prompt did not define calculation logic clearly.	Ask Kimi to explain every formula, then regenerate only the affected sheet.
Spreadsheet has static values instead of formulas	The prompt asked for a result table, not a calculated workbook.	Ask for formulas instead of hardcoded totals and request a Verification sheet.
Data is missing	Source file had hidden filters, unreadable PDF sections, or unclear extraction rules.	Compare row counts and ask Kimi to list skipped or uncertain records.
Duplicate rows appear	Multiple files contained overlapping records.	Ask Kimi to deduplicate by a specific key such as invoice ID, email, SKU, or company name.
Charts do not match the data	Chart source range may point to the wrong table or partial range.	Inspect chart source range and ask for a chart-source note in the workbook.
Pivot table is wrong	Wrong source range, aggregation, grouping, or filter.	Check pivot settings and ask Kimi to document pivot source and aggregation logic.
Web-sourced data is outdated	The task did not specify a date range or source quality.	Ask Kimi to cite sources, include retrieval dates, and prioritize official or recent sources.
Assumptions are not marked	The prompt did not require assumption labeling.	Ask Kimi to create an Assumptions sheet and move every uncertain value there.
XLSX is not available	You may be in normal chat or the task may not have generated a workbook output.	Use the Sheets shortcut or Agent path, ask for an XLSX workbook, and preview before download.

When Kimi Sheets is not enough

Kimi Sheets can speed up spreadsheet creation, but it should not be the final authority for high-stakes work.

Use a qualified accountant or financial analyst for financial models, valuation, taxes, forecasts, and investor materials.
Use a legal professional for contract, compliance, regulatory, or litigation spreadsheets.
Use a data analyst for large datasets, complex statistical work, or production dashboards.
Use academic review for research matrices, citations, experimental data, and grading-related spreadsheets.
Use Excel, Google Sheets, Power Query, Power Pivot, BI tools, or databases when you need repeatable refresh pipelines, macros, live collaboration, or governed analytics.

Final checklist before you trust the spreadsheet

Open the downloaded XLSX file locally.
Save a clean copy before editing.
Check all expected sheets are present.
Compare source row counts with generated row counts.
Check duplicate rows and missing required fields.
Inspect formulas in calculated columns.
Manually recalculate sample rows.
Confirm totals, subtotals, percentages, and growth rates.
Trace cross-sheet references.
Check pivot-table source ranges and aggregation logic.
Check chart source ranges, axes, legends, labels, and filters.
Review assumptions and move hidden assumptions into a dedicated sheet.
Open source URLs and verify web-sourced numbers.
Check date, currency, number, and percentage formats.
Get expert review for finance, legal, medical, academic, compliance, or business-critical spreadsheets.

Sources checked

The product and workflow sections retain these official sources from the October 12, 2026 review: the Kimi Sheets product page, the Kimi Docs & Sheets Help Center overview, the Kimi Sheets use cases and prompt library, the Kimi capabilities and supported files FAQ, the Kimi credit update and usage rules, and the Kimi data usage and sharing page.

The October 15, 2026 privacy and workspace update checked the Kimi Business workspace documentation, Business FAQ, Business platform-support guidance, kimi.ai Privacy Policy, and kimi.com data-usage guidance. The support-based opt-out process above is attributed to the kimi.com help page, not presented as one global procedure for every Kimi account.

FAQ

What is Kimi Sheets?

Kimi Sheets is Kimi’s AI spreadsheet agent for creating, editing, cleaning, analyzing, and exporting Excel-style workbooks. It can work from natural-language instructions and uploaded source files.

How do I create a spreadsheet with Kimi Sheets?

For work data, first confirm authorization and select the approved workspace. Then use Kimi’s Agent button or Sheets shortcut, describe the task, upload approved source files, review the planned sheets and formulas, preview the result, and download the XLSX. When Business controls are required, use a supported web or desktop Enterprise Workspace.

Can Kimi Sheets create XLSX files?

Yes. Kimi’s Help Center says Kimi Sheets can output XLSX files directly on the web, and users can preview the spreadsheet structure online before downloading the file.

Can I upload Excel or CSV files to Kimi Sheets?

Yes. Kimi’s Help Center says users can upload an existing Excel or CSV file and explain what they want done. Kimi’s broader file FAQ also lists common file formats such as PDF, Word, Excel, PPT, images, TXT, and video.

Can Kimi Sheets create formulas, pivot tables, and charts?

Kimi’s product page describes formula generation, pivot tables, charts, dashboards, multi-sheet logic, and spreadsheet formatting. You should still verify formulas, pivot source ranges, chart ranges, filters, and assumptions manually after downloading the XLSX file.

How do I verify a spreadsheet created by Kimi Sheets?

Open the XLSX file locally, check row counts, duplicates, missing values, date and currency formats, formulas, cross-sheet links, pivot-table source ranges, chart source ranges, web sources, and assumptions. Manually recalculate sample rows before trusting the workbook.

Can Kimi Sheets build financial models?

Kimi’s use-case library includes DCF valuation and financial modeling examples. However, financial models should be reviewed by a qualified person because assumptions, formulas, discount rates, and source data can materially change the result.

Does Kimi Sheets use credits?

Yes. Kimi’s membership documentation says Sheets is one of the features that share a single credit pool with Agent, Deep Research, Slides, Docs, Kimi Code, Kimi Work, and other membership features. Usage depends on actual task consumption, including tokens, model choice, context length, and tools.

Is Kimi Sheets safe for confidential data?

It depends on the account, workspace, agreement, and data. Obtain organizational approval before uploading. Kimi Business documents a no-training default for enterprise content; a personal subscription is not the same coverage. For personal accounts, check the applicable policy and whether any requested training opt-out has taken effect. The mobile app does not currently support Business, so use a supported web or desktop enterprise session when those controls are required. No training does not guarantee zero retention or immediate deletion. See the privacy and workspace checks.

Can Kimi Sheets replace Excel?

No. Kimi Sheets can help create and structure Excel-style workbooks, but you still need spreadsheet software such as Excel, WPS, or a compatible editor to inspect formulas, verify charts, adjust formatting, and maintain the file over time.

When should I not use Kimi Sheets as the final authority?

Do not use Kimi Sheets as the final authority for finance, legal, medical, academic, compliance, tax, payroll, investment, or business-critical spreadsheets. Use it for drafting and structuring, then verify the workbook manually or with a qualified expert.