JevMade Sign in
← Back to guides

JevMade field notes / Integration guide

A guide to preparing an AI assistant for incoming tasks

This guide explains how to use software that helps an AI assistant pick the right AI service level, executable tools, and saved instructions for a task before it starts working.

Original by JoacoMarcAgent workflows

Listen to this guide

JevMade’s plain-English explanation

0:00 /

Our summary

This project is a sorting tool for an AI assistant. When a user makes a request, the software decides how difficult the task is. It then chooses the right AI service level, executable tools, and saved instructions needed to complete the job without wasting time or money.

The setup uses Jev, an AI tool that chooses from a list of options instead of writing out an answer. It makes one fast request to pick the best setup from your saved lists. If the network takes too long, it falls back to simple keyword matching.

This guide is for people building AI software who want to separate task sorting from task execution. The tool requires a private access code to work. The actual financial costs and available AI services depend entirely on the external accounts the reader chooses to connect.

Key takeaways

  1. List your AI services from cheapest to most capable so the software can escalate when needed.
  2. Separate the step that decides how to handle a task from the step that actually does it.
  3. Test the sorting rules with the provided evaluation commands before changing how the software makes choices.

The success rates and response times reported in the guide happened in the author's specific tests on 181 examples. Actual performance and financial costs will vary based on the user's own setup and location.

Repository README

Read the original guide Opens the author’s site in a new tab.

Keep this for later

Sign in to bookmark experiments, guides and videos, and keep notes only you can see.

Continue to sign in

We’ll bring you back to this listing.