JevMade field notes / Worked recipe

Sort support messages with four separate questions

This recipe asks Jev which team a message needs, whether it is urgent, how frustrated the customer sounds, and whether they want a refund.

Original by Jeroen Erne / NexibeoClassification

Listen to this guide

JevMade’s plain-English explanation

0:00 /

AI narration

Credits and license

“01 · Support triage” 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

A customer message can need several decisions at once. Jeroen Erne's recipe uses Jev, an AI tool that chooses and rates answers, to identify the right team, urgency, frustration, and a request for money back. Each question addresses a different part of the message.

The program sends all four questions together. Descriptions explain where one team's work ends and another's begins. The program then uses the answer's confidence to send the message to a team, ask a person to check it, or hold it for later review.

The author reports results from a small set of hand-labelled tickets, including ambiguous messages. Those results do not establish accuracy for another company's inbox. Teams need their own examples and rules; a refund flag identifies a request rather than granting money back.

Key takeaways

Describe the boundaries between teams, not only their names.
Keep urgency and frustration as separate questions.
Test the review cutoffs on messages your own team finds ambiguous.

More guides →See the experiments →