Data Flow Overview

The pipeline

1. Your data arrives — you upload a document, or your accounting software sends invoices across automatically.
2. We read it — the platform pulls out each line of spending.
3. We sort it — fixed rules decide what each line is. Only if no rule fits do we ask an AI for a suggestion, and personal details are stripped out before anything is sent.
4. You approve it — someone at your company checks each line. We only calculate from lines you have approved.
5. We calculate it — each approved line item gets matched to an emission factor, using our tiered method, and turned into a CO₂e figure (see the Emission Factor Dictionary).
6. We store it — every result stays linked to the document it came from, so any number can be traced back to its evidence.
7. You publish it — results start as drafts. They only count towards your reported footprint once you publish them.
8. It gets deleted — by you, by our automatic clean-up, or by erasing your whole account. Deletion removes everything connected to it (see Data Deletion).

Stage detail

1. Your data arrives. There are two routes in. You can upload an invoice or supporting document yourself, or you can connect your accounting software (Xero, QuickBooks or Sage) and let it send invoices and bills across on its own. Either way, the original file is stored securely and joins the queue for processing.

2. We read it. PDFs and images are scanned to pull the text out. Spreadsheets are read directly. Either way, the result is a list of spending lines, ready to be sorted.

3. We sort it. Each line is matched against fixed rules first: your own account codes, corrections you have made for a supplier before, built-in patterns for well-known suppliers, and commodity codes. Only when none of those fit does an AI get involved — and before any text is sent to it, personal and financial details are automatically detected and stripped out. Even then, the AI only *suggests* a category. It never sets one itself.

4. You approve it. Someone at your company confirms or corrects every line. We calculate only from approved lines, so your figures always come from something a person has actually checked, never from automatic processing alone.

5. We calculate it. For each approved line item, we work out the right emission factor using our tiered method — supplier-declared Gold, activity-based Silver, or spend-based Bronze — and turn it into kilograms of CO₂e. Only the numbers and the category go to our emission factor provider. No personal data is sent.

6. We store it. Each result stays linked to the spending line and the document it came from, so you can always trace a number back to the evidence behind it. Stored files are encrypted, and your company's data is kept separate from every other customer's.

7. You publish it. New emission records start as drafts. Publishing is a deliberate action, and only published figures show up in your dashboard, your reports and your public profile.

8. It gets deleted. Data leaves in one of three ways: you delete a document, our automatic clean-up removes something that has reached the end of its 7-year holding period, or you ask us to erase your account entirely. Deletion is permanent and takes everything connected with it, across both our storage and our database. If you ask for something to be deleted, it goes immediately rather than waiting out the 7 years — apart from the small set of records the law requires us to keep.