JevMade field notes / Architecture and workflow guide

How to build a fast voice assistant for Mac computers

This guide explains how to make a voice-controlled software assistant. It shows how the program quickly picks actions from a list, handles complex screen tasks, and avoids getting stuck in endless loops.

Original by Kevin BahrabadiAgent workflows

Listen to this guide

JevMade’s plain-English explanation

0:00 /

AI narration

Credits and license

“Jev Voice” by Kevin Bahrabadi. 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 Kevin Bahrabadi

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 is a voice-controlled software assistant for Mac computers. People use it to open applications, search the internet, or type text without using their hands. It uses Jev, an AI tool that chooses from a list of options instead of writing answers.

When you speak, the software turns your voice into text. Instead of figuring out steps one by one, it asks the AI several questions at once. The AI picks the best action from a safe list, and the computer carries it out.

This guide helps software builders make voice assistants faster and safer. The program stops automatically if it gets stuck repeating the same failed action. A person can watch the software's choices on a separate screen and pause the process if needed.

Key takeaways

Ask the AI multiple questions at the same time to speed up its decision process.
Reuse an AI decision only if the screen has not changed since the last action.
Show the AI's choices and certainty levels on a screen so a person can check them.

More guides →See the experiments →