JevMade field notes / Worked recipe

Check duplicate records before merging them

This recipe compares pairs of organization records and separates likely matches, different organizations, and cases that need a person's judgment.

Original by Jeroen Erne / NexibeoClassification

Bookmark Keep this in your collection.

Leave a noteWhat would you try with this?

Only you can see your notes.

Listen to this guide

JevMade’s plain-English explanation

0:00 /

AI narration

Independent ElevenLabs affiliate. JevMade may earn a commission.

Credits and license

“06 · Duplicate detection” by Jeroen Erne / Nexibeo. Read the original source.

This expanded guide is an AI-narrated adaptation of the source’s essential explanation, examples and caveats, not a word-for-word reading. The synthetic voice does not imitate the author or imply their endorsement.

License: MIT

MIT License

Copyright (c) 2026 Jeroen Erne (https://nexibeo.com, https://completeaitraining.com)

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.

Our summary

Importing records from different systems can create duplicates with different spellings or abbreviations. This recipe asks Jev whether two records describe the same organization at the same location. Its policy keeps separate branches and parent companies distinct rather than merging everything with a similar name.

Ordinary software cleans up website addresses and phone numbers and supplies the comparison results as facts. Jev rates the remaining evidence, and the program maps that rating to keep, review, or merge. The author compares this approach with showing only the raw records.

Small score changes can cross a sharp cutoff, so uncertain pairs need a genuine review range. Candidate selection also matters: a larger system should cheaply shortlist plausible pairs first. The author's small sample is not evidence that automatic merging is safe for another organization's data.

Key takeaways

Define whether branches and subsidiaries count as separate organizations.
Compute exact comparisons before asking the AI to judge.
Leave uncertain matches for review before changing records.

More guides →See the experiments →