INVOICE PROCESSING

Fingerprint S3-dropped invoices against the paid ledger before approval

Watches an S3 bucket for new invoice PDFs, computes a content fingerprint (vendor + amount + invoice number + date).

CategoryInvoice Processing
Enginesim
Difficultyintermediate
Triggerevent
Steps6
Setup~15 min

How it runs

The automated pipeline, trigger to output.

  • TriggerNew invoice PDF lands in S3 intake bucketAWS S3
  • ActionParse PDF and extract vendor, number, amount, date
  • ActionCompute normalized fingerprint hash
  • ActionQuery Postgres paid-ledger for matching fingerprintPostgreSQLPostgres
  • LogicBranch: duplicate match vs. clean
  • OutputQuarantine alert to Slack or write to approval queueSlack

What it does

This workflow stops duplicate payments at the front door. Every invoice that lands in your intake S3 bucket is parsed, normalized, and turned into a deterministic fingerprint. That fingerprint is matched against your already-paid ledger in Postgres. Clean invoices route to approval; suspected duplicates are quarantined and the AP team is alerted in Slack with the matching prior payment.

When to use it

Use it when vendors email or re-upload the same invoice twice, when multiple intake channels feed one bucket, or when you've been bitten by paying the same invoice number under a slightly different file name. It runs entirely before approval, so no human ever sees a confirmed duplicate.

How it works

  1. 1A new object in the S3 invoice-intake bucket triggers the run.
  2. 2The PDF is parsed and key fields are extracted (vendor, invoice number, amount, date).
  3. 3A normalized fingerprint hash is computed from those fields.
  4. 4Postgres is queried for any paid invoice with a matching fingerprint.
  5. 5A logic branch splits on match found vs. clean.
  6. 6Duplicates post to a Slack quarantine channel with the prior payment reference; clean invoices are written back to the pending-approval queue.

Set it up

What you configure once, before turning it on.

  1. 1
    Connect AWS S3Buckets, objects, signed URLs.
  2. 2
    Connect PostgresAny Postgres URL — query, write, migrate.
  3. 3
    Connect SlackChannels, DMs, threads, mentions.
  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.