How to use AI to reformat an hours import file in Every

Last updated: July 9, 2026

Every pay period, you can use an AI assistant to map your timesheet data into Every's import format in seconds. No manual copying, no reformatting.

This works with any timesheet export — whether your hours are in a totals column or spread across daily columns, formatted as "8h 30m", "8:30", or "8.5", and whether your employees are identified by name or email.

What you'll need

  • Your Every import template. Download it from the Import hours page inside your payroll cycle.

  • Your timesheet export from your time tracking tool (Excel, CSV, or any spreadsheet format).

Step 1 — Open your AI assistant

Go to claude.ai, chatgpt.com, or whichever AI tool you use. Start a new conversation.

Step 2 — Upload both files

Attach both files to the same message before sending:

  1. Your Every import template (the CSV you downloaded)

  2. Your timesheet export

Step 3 — Send this prompt, and it will return the Every import csv filled out, along with any issues it found

Copy and paste the following exactly:

I have two files attached. The first is an Every.io hourly hours import template. The second is a timesheet export from our time tracking tool.

Your goal: fill in the hours column of the Every template using data from the timesheet. Follow these rules:

Step 1 — Understand the Every template
Identify the columns available. The template will have an hours column, and at least one of the following to match employees: employee_name, employee_email, or employee_id. Note which identifier columns are present.

Step 2 — Understand the timesheet
Examine the timesheet and figure out its structure on your own. Specifically:





Find where employee names or emails appear.



Determine how hours are recorded. Look for a pre-calculated totals column first. If there is no totals column, sum the individual day or shift columns yourself.



Identify the time format being used and convert all hours to a plain decimal number rounded to two decimal places. Handle any format you encounter, including but not limited to: 14h 12m, 14:12, 14.20, 8.5, 0h 30m, 1:30:00, or plain integers.

Step 3 — Match employees
Match each employee in the timesheet to a row in the Every template using whatever identifier is available: name, email, or both. Be flexible with name formatting (ALL CAPS, first name only, nicknames, trailing spaces). If you find a match you are not confident about, flag it rather than guessing silently.

Step 4 — Fill the template





Fill in the hours column for each matched employee.



Do not modify employee_id, employee_name, employee_email, or any other column. Only fill hours.



Only include employees with hours greater than zero. Skip anyone with a blank, dash, or zero total.



If an employee appears in the Every template but not in the timesheet, leave their row blank and flag them.



If an employee appears in the timesheet but not in the Every template, flag them but do not add a new row.

Step 5 — Return the result
Return the completed CSV ready to download. Then provide four clearly labeled lists:

Matched and filled — employees who were matched and had hours greater than zero. Show name and hours.

Flagged matches — employees where you matched but were not fully confident. Show the timesheet name, what you matched it to in the Every template, and why you are uncertain.

Not matched — employees who appear in the timesheet but could not be found in the Every template at all.

Zero or no hours — employees who appear in the timesheet but had a blank, dash, or zero total. List them explicitly so you can confirm they truly worked no hours this period and are not a data entry gap.



Step 4 — Review before uploading

Check the four lists before downloading the CSV:





Confirm all flagged matches look right. If a name was matched incorrectly, tell the AI: "CALEB R. in the timesheet is Caleb Rhinehart in the Every template — please fix."



Review the zero-hours list. If someone should have hours, go back to the timesheet and check.



Make sure no employee IDs were changed.

Once everything looks right, download the completed CSV.



Step 5 — Upload to Every

Go back to the Import hours page in your payroll cycle and upload the file. Every will validate it and show you a preview before anything is applied to payroll.