Complete AI Training
Sign inGet my AI kit

Your job's AI kit

Get your AI kit

Tell us who you are and what you do. We show you your kit right away and email you the link: skills, prompts, AI agents, MCP servers and courses for your job.

500+ jobs ready, and we make a kit for any other job. No payment needed to look.

Share

AI agent for data entry specialists

Duplicate Customer Record Merge Agent

Duplicate records merged safely, with all linked history preserved and a trail of what changed

Data cleanupIT and developmentOperations
See how it worksGet access
Duplicate Customer Record Merge Agent: what goes in, what the agent does and what you get

What it does

Two records for the same customer split orders and invoices across both, and merging them by hand can lose history. This agent searches for likely duplicates using name, address, phone and email, and scores each pair by how well these match. For each pair it reads linked transactions, notes and open items and compares them. It drafts a merged record that keeps the best value for each field and a list of what moves where. Before presenting a pair it checks that no linked order, invoice or leave record would be lost or attached to the wrong party. A person approves each merge before it is applied. Edge case: two employees with the same name and birthday but different national numbers are marked as not duplicates.

How it works

Follow the arrows from top to bottom. The orange dashed arrow is the loop: when a check fails, the agent goes back and tries again.

Start and resultWhat it doesA check on its own workWaits for your OKGoes back and retries
Yes, continueApprovedYes, continueNoNo 1 STARTS WHEN Weekly duplicate scan 2 USES A TOOL Find candidate pairs by name, address, phone andemail 3 DOES Score each pair and discard those with conflictingID fields 4 USES A TOOL Read linked transactions, notes and open items foreach pair 5 DOES Draft the merged record and the list of items tomove 6 CHECKS THE RESULT Would every linked transaction and history item bekept? If not: fix the move list or drop the pair from thisrun. Back to step 4. 7 YOU APPROVE Specialist approves each merge 8 USES A TOOL Apply the approved merge 9 CHECKS THE RESULT Do totals and counts match before and after? If not: stop and restore the original records from thebackup. Back to step 7. 10 RESULT Merge log with before and after values
Read the steps as a list
  1. Weekly duplicate scan
  2. Find candidate pairs by name, address, phone and email
  3. Score each pair and discard those with conflicting ID fields
  4. Read linked transactions, notes and open items for each pair
  5. Draft the merged record and the list of items to move
  6. Would every linked transaction and history item be kept?If not: fix the move list or drop the pair from this run. Back to step 4.
  7. Specialist approves each mergeThe agent waits here for your OK.
  8. Apply the approved merge
  9. Do totals and counts match before and after?If not: stop and restore the original records from the backup. Back to step 7.
  10. Merge log with before and after values

How it decides

A pair is proposed when its match score is above the threshold and no strong field conflicts, such as different ID numbers, is found.

  • Propose a pair when the score is 85 or higher
  • Never propose a pair with different tax or national IDs
  • Keep the most recent verified value for contact fields
  • Send pairs scoring 70 to 84 to a manual review list

Make it yours

Every agent is a starting point. You choose these settings for your own situation.

  • Match score to propose a merge (default 85)
  • Fields that must not conflict
  • How to pick the surviving record
  • Review list threshold (default 70)

What keeps you in control

It always asks you first

  • Each merge before it is applied

Hard limits

  • Never merges without approval
  • Always keeps a restorable copy of both records

It stops when

  • Done: all approved merges verified
  • Stop: a before and after count does not match and cannot be restored

Set it up

We guide you through the set-up, step by step

Members get the full set-up guide for this agent. No technical skills needed: you copy, paste and upload.

10 minto set it up in your AI
5 AIsChatGPT, Claude, Copilot, Gemini, Grok
  • One set of instructions to paste into your AI, with the clicks for ChatGPT, Claude, Microsoft 365 Copilot, Gemini and Grok
  • The agent then walks you through connecting your own data, one source at a time
  • A downloadable copy with the flow chart, the rules and the full guide
Get access to this agent

An example run

What happensThe scan found 38 possible pairs. A pair for Ann Marlow and A. Marlow scored 91, with matching phone and address. The first check failed because 2 invoices were tied to a second billing account. The agent added them to the move list. After approval it merged, and the check showed 14 orders and 9 invoices on both sides. Another pair with different tax IDs was rejected.

More agents for data entry specialists

Quality control · Data cleanup

Insurance Bordereaux Validation Agent

Bordereaux that pass schema and reconcile to control totals.

7 steps1 check2 jobs
Data cleanup · Operations

Inventory Master Data Agent

A clean item master with one record per real item.

7 steps1 check1 job
Compliance · Administration

Visitor Log and Compliance Audit Agent

Keep the visitor log complete and audit-ready with evidence for every correction.

11 steps2 checks2 jobs
Quality control · Science and research

Property Database Entry Check Agent

Allow only complete, consistent and plausible property entries into the database.

11 steps2 checks3 jobs