INVOICE PROCESSING

Agentic Vendor Entity Resolution for Duplicate Holds

When a possible duplicate is flagged, an agent investigates whether two differently-named vendors are actually the same entity, gathers evidence, and recommends a hold-or-pay…

CategoryInvoice Processing
Enginepaperclip
Difficultyadvanced
Triggerwebhook
Steps5
Setup~25 min

How it runs

The automated pipeline, trigger to output.

  • TriggerCandidate duplicate pair receivedHTTP webhook
  • ActionPull both vendor masters and historiesSnowflakeSnowflake
  • LogicEstimate same-entity probability from shared signals
  • ActionConfirm parent/subsidiary via web searchExa
  • OutputPost hold-or-pay brief to approverSlack

What it does

The hardest duplicates hide behind vendor identity: "Acme LLC" and "Acme Holdings Inc" billing the same service. When a candidate pair is flagged, this agent investigates whether the two vendors are really one entity. It pulls both vendor records and payment histories from Snowflake, weighs signals like shared bank details, addresses, and tax IDs, optionally checks public records, and produces a hold-or-pay recommendation with a written rationale routed to the approver.

When to use it

Use it for the gray-area cases your deterministic scanners cannot resolve alone, where a human would otherwise spend twenty minutes cross-referencing vendor masters. It turns ambiguous flags into a decision-ready brief.

How it works

  1. 1A candidate duplicate pair arriving via webhook triggers the agent.
  2. 2The agent queries Snowflake for both vendors' master records and payment histories.
  3. 3It reasons over shared bank, address, and tax-ID signals to estimate same-entity probability.
  4. 4It optionally runs a web search to confirm parent-subsidiary or name-change relationships.
  5. 5It writes a hold-or-pay recommendation with evidence and posts the brief to the approver in Slack.

Set it up

What you configure once, before turning it on.

  1. 1
    Connect HTTP webhookTrigger any URL on agent actions.
  2. 2
    Connect SnowflakeWarehouses, queries, shares.
  3. 3
    Connect ExaNeural search across the web.
  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.