INVOICE PROCESSING

Duplicate and suspicious vendor invoice screen from Outlook

Screens every emailed vendor invoice against historical invoices in BigQuery to catch duplicates and suspicious bank-detail changes.

CategoryInvoice Processing
Enginesim
Difficultyintermediate
Triggerevent
Steps6
Setup~15 min

How it runs

The automated pipeline, trigger to output.

  • TriggerVendor invoice arrives in Outlook AP inboxOutlook
  • ActionExtract vendor, invoice number, amount, bank details with OpenAIOpenAI
  • ActionQuery vendor invoice history and known bank details in BigQueryGoogle BigQueryBigQuery
  • LogicFlag duplicates and changed remit-to bank accounts
  • ActionRecord cleared invoices in BigQueryGoogle BigQueryBigQuery
  • OutputAlert AP lead on Slack for duplicates or fraud riskSlack

What it does

Before an invoice ever reaches matching, this flow checks it for the two cheapest ways AP loses money: paying the same invoice twice and paying a fraudulently altered bank account. It fingerprints each new invoice and compares it to the vendor's history.

When to use it

Run this as a front gate on your AP intake when you process invoices from many vendors and worry about duplicate payments or payment-redirection fraud. It is a guardrail, not a full matcher, so it pairs well with a separate PO-match flow.

How it works

  1. 1A vendor invoice arrives in the Outlook AP inbox.
  2. 2OpenAI extracts vendor name, invoice number, amount, and the remit-to bank details.
  3. 3BigQuery is queried for prior invoices from the same vendor and for the vendor's last-known banking details.
  4. 4Logic flags a duplicate if invoice number or amount-plus-date already exists, and flags fraud risk if bank details changed.
  5. 5Clean invoices are recorded in BigQuery as cleared for processing.
  6. 6Any duplicate or changed-bank invoice triggers a Slack alert tagging the AP lead with the specific reason.

Set it up

What you configure once, before turning it on.

  1. 1
    Connect OutlookMail, calendar, contacts.
  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.