INVOICE PROCESSING

Validate invoice math and tax before posting to the ledger

OCRs each invoice, recomputes line subtotals, tax, and grand total, and either posts arithmetically sound invoices to Postgres or returns math errors and tax mismatches…

CategoryInvoice Processing
Enginesim
Difficultyintermediate
Triggerevent
Steps5
Setup~15 min

How it runs

The automated pipeline, trigger to output.

  • TriggerNew invoice emailGmailGmail
  • ActionOCR line items and stated totalsOpenAI
  • LogicRecompute subtotal, tax, and grand total
  • ActionPost validated invoice to ledgerPostgreSQLPostgres
  • OutputEmail math or tax discrepancy to vendorGmailGmail

What it does

Guards your ledger against arithmetic and tax errors. The flow extracts every line item, independently recomputes the subtotal, applies the expected tax rate for the vendor's jurisdiction, and compares its computed grand total to the printed total. Clean invoices post; failing ones bounce back to the vendor with the exact discrepancy.

When to use it

Use this when invoices frequently arrive with rounding errors, wrong tax rates, or totals that don't foot, and you want to catch them before they enter the books rather than during reconciliation.

How it works

  1. 1A new invoice email triggers the run.
  2. 2OCR extracts line items, printed subtotal, tax, and grand total.
  3. 3The flow recomputes subtotal, applies the expected tax rate, and derives its own total.
  4. 4A validation check compares computed values to printed ones within a cent tolerance.
  5. 5Passing invoices are posted to the Postgres ledger as ready for approval.
  6. 6Failing invoices trigger a reply email to the vendor itemizing the math or tax mismatch.

Set it up

What you configure once, before turning it on.

  1. 1
    Connect GmailRead, draft, send, label.
  2. 2
    Connect OpenAIModels, embeddings, files.
  3. 3
    Connect PostgresAny Postgres URL — query, write, migrate.
  4. 4
    Set each agent's modelWe leave models unset so you pick the tier — fast + cheap, or top-quality.
  5. 5
    Tune it to your dataEdit the prompts, filters, and field mappings so it matches how your team works.
  6. 6
    Test, then turn it onRun once against a sample, confirm the output, then enable the trigger.

Run this workflow in your colony.

14-day trial. No DevOps. No Sales call. Provisioned in under a minute.