INVOICE PROCESSING

Three-Way Invoice Pre-Match from Dropbox to BigQuery

Watches a Dropbox folder for new vendor invoice PDFs, extracts each line item, matches it against PO and goods-receipt records in BigQuery.

CategoryInvoice Processing
Enginesim
Difficultyintermediate
Triggerevent
Steps6
Setup~15 min

How it runs

The automated pipeline, trigger to output.

  • TriggerNew invoice PDF added to Dropbox folderDropboxDropbox
  • ActionExtract invoice line items with OpenAIOpenAI
  • ActionQuery PO and goods-receipt lines in BigQueryGoogle BigQueryBigQuery
  • LogicCompare qty and price per line within tolerance
  • ActionWrite approved-to-pay row to BigQueryGoogle BigQueryBigQuery
  • OutputPost mismatch details to Slack AP channelSlack

What it does

Automates the classic accounts-payable three-way match. When a vendor drops an invoice into a shared Dropbox folder, the workflow reads every line, compares quantity and unit price against the matching purchase order and the receiving record in BigQuery, and decides whether the invoice can move straight to approval or needs a human to investigate.

When to use it

Use it when your AP team manually opens PDFs and cross-checks them against the ERP. Best for teams whose PO and receiving data already lands in BigQuery and who want a fast, auditable pre-match before anything hits the approval queue.

How it works

  1. 1A new PDF in the Dropbox invoices folder triggers the run.
  2. 2OpenAI parses the invoice into structured lines (PO number, SKU, quantity, unit price, totals).
  3. 3BigQuery is queried for the referenced PO lines and their posted goods receipts.
  4. 4A match-logic step compares quantity and price per line within tolerance and classifies the invoice as matched or exception.
  5. 5Matched invoices write an "approved-to-pay" row back to BigQuery.
  6. 6Exceptions post to a Slack AP channel with the specific mismatched lines and dollar variance.

Set it up

What you configure once, before turning it on.

  1. 1
    Connect DropboxFiles and folders.
  2. 2
    Connect OpenAIModels, embeddings, files.
  3. 3
    Connect BigQueryDatasets, queries, schemas.
  4. 4
    Connect SlackChannels, DMs, threads, mentions.
  5. 5
    Set each agent's modelWe leave models unset so you pick the tier — fast + cheap, or top-quality.
  6. 6
    Tune it to your dataEdit the prompts, filters, and field mappings so it matches how your team works.
  7. 7
    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.