INVOICE PROCESSING

Email Invoice Ingest, Extract, and Three-Way Match

Watches a shared AP mailbox, extracts line items from invoice PDFs with AI, matches them against PO and receipt data in Postgres.

CategoryInvoice Processing
EngineSim + Paperclip
Difficultyadvanced
Triggerevent
Steps6
Setup~25 min

How it runs

The automated pipeline, trigger to output.

  • TriggerNew invoice email in AP mailboxGmailGmail
  • ActionExtract line items and PO number from PDF with AIOpenAI
  • ActionFetch matching PO and receipt rows from PostgresPostgreSQLPostgres
  • LogicThree-way compare totals and quantities
  • ActionWrite passing invoices to payment queuePostgreSQLPostgres
  • OutputEmail vendor the mismatched lines on failureGmailGmail

What it does

Turns a noisy AP inbox into a structured match pipeline. It reads invoice emails, extracts vendor, PO number, and line items from the attached PDF, then runs a three-way match against your procurement tables in Postgres. Clean invoices are queued for payment; failures trigger an automatic, polite clarification email to the vendor.

When to use it

Use this when invoices arrive as email attachments rather than through a portal, and someone is keying them in by hand before matching. Best for mid-volume AP teams that already keep PO and receipt data in a database.

How it works

  1. 1A new email arrives in the monitored AP Gmail mailbox (trigger).
  2. 2The PDF attachment is sent to an OpenAI extraction step that returns structured line items and the PO number.
  3. 3A query pulls the matching PO and goods-receipt rows from Postgres.
  4. 4A logic step performs the three-way comparison on totals and quantities.
  5. 5Passing invoices are written to the payment queue table in Postgres.
  6. 6Failing invoices generate a vendor reply via Gmail listing the exact mismatched lines and what is needed to resolve them.

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.