SpreadsheetAutomation
Take the macro nobody dares touch and the weekly merge-and-clean routine, and rebuild them as monitored, documented automations.
Spreadsheet Automation: what the work involves
Every company runs on a few spreadsheets that one person understands. Files are emailed around, versions multiply, a hidden column breaks a formula, and the monthly consolidation takes an afternoon of careful copy and paste. When that person is on leave the process stops, and when the numbers look odd nobody can tell which step introduced the mistake.
DevKey audits the workbook, writes down what it really does, and rebuilds the repeatable parts in code that runs on a schedule or on demand. Files are collected, cleaned, merged and checked, with the results written back to Google Sheets, Excel or a database. A language model is added only where judgement over messy text is needed, such as tidying free-text categories or drafting a note about unusual rows. Every run leaves a log, so errors are traceable.
Core features
Workbook audit and documentation
We trace formulas, macros and manual steps, and hand back a plain description of what the sheet does and where it can fail.
Scheduled consolidation
Files from many people or folders are fetched, aligned to one layout and merged automatically, with a clear message about any file that does not fit.
Data cleaning rules
Dates, phone numbers, names and categories are standardised, duplicates are flagged, and the original value is preserved beside the cleaned one.
Natural-language helpers
Staff can ask for a pivot, a filter or a formula in plain words, and the helper shows the steps it will apply before touching the data.
Validation checks
Totals are reconciled between input and output, and unexpected blanks or outliers stop the run with an explanation rather than flowing into a report.
Move-to-database path
When a sheet has outgrown itself, we migrate the core tables to a database and keep a familiar spreadsheet view on top.
What we get right before launch
Hidden logic in old workbooks
Legacy sheets carry undocumented rules, hard-coded values and manual overrides. We reproduce results on historical months and investigate every difference before declaring the new routine correct.
Sensitive data in shared files
Payroll, customer and pricing sheets travel through email and chat. We lock permissions, avoid sending whole sheets to external models, and mask columns that a language model does not need to see.
Spreadsheets are not always the answer
Sometimes the honest recommendation is a small application or a proper database. We say so when concurrency, audit or size make a sheet the wrong foundation, even if it means a smaller engagement.
Tools and technology
- Python
- pandas
- Google Apps Script
- Excel Office Scripts
- OpenAI GPT
- PostgreSQL
- n8n
- FastAPI
Common questions, answered
Can you fix our existing macro instead of replacing it?
Sometimes. If the macro is small and stable we document and harden it. If it depends on one desktop and manual clicks, rebuilding in a scheduled script is usually safer and easier to monitor.
Does it work with Google Sheets and Excel both?
Yes. We read and write both through their official interfaces, and can sync between them. Where a team needs both, we agree which one is the source of truth to avoid conflicting edits.
Will AI change our numbers?
Numbers are calculated by code with checks against the source. A model is limited to text tasks such as classifying descriptions or summarising notes, and its outputs are labelled so reviewers can tell them apart.
How do we get started?
Send us the workbook and a description of the monthly routine. We review it, list the manual steps and risks, and propose which parts to automate first, usually the one that costs the most staff time.
More AI Agents & Automation services
All AI Agents & Automation servicesData Entry Automation
Stop paying skilled staff to retype information from one screen into another; capture it once, validate it, and write it to the right system.
Automated Report Generation
Have weekly and monthly reports assemble themselves from live data, complete with charts and a plain-language summary ready for a quick review.
Chat With Your Data Solutions
Give managers a chat box that answers questions from your real business data, shows its working, and refuses when it cannot be sure.
Ready to start your Spreadsheet Automation project?
Tell us what you need and we will come back with a clear scope, timeline and the questions worth answering before any build starts.