JevMade field notes / Runnable examples

How to sort customer messages and help articles using AI choices

This guide explains how to use an AI tool to organize customer support messages and find helpful articles. It shows how to ask the software questions and set rules for when to trust its choices.

Original by Rajiv KuriakoseIntegrations

Listen to this guide

JevMade’s plain-English explanation

0:00 /

AI narration

Credits and license

“Worked examples via OpenRouter” by Rajiv Kuriakose. 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 Rajiv

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

This project provides working examples for sorting customer support messages and finding helpful articles. It uses Jev, an AI tool that chooses from specific options rather than writing an answer. People use this setup to automatically route incoming questions.

The software asks the AI multiple questions about a single message at once. To find the best help article, it instead sends many articles at the same time and asks which one answers the customer. Normal software rules handle escalations.

This guide helps people building automated support systems. The examples show how to structure requests and handle uncertain answers before the software acts. The author provides these examples to demonstrate the setup rather than to prove overall performance.

Key takeaways

Group questions together for a single message, but check different articles using separate simultaneous requests.
Require higher certainty from the AI before taking serious actions like issuing a refund.
Report ties accurately when comparing search methods, as assigning a random rank makes comparisons misleading.

More guides →See the experiments →