Practical AI at Work: Create Your Own AI Work Companion

Live hands-on webinar

Watching prompts run across several Codex chats can get boring, and it is easy to forget which chat is working, waiting for input, ready for review, or has run into a problem.

What if a little animated pet showed you those states at a glance while making your workspace feel more like your own?

In this live one-hour Practical AI at Work session, Bill Raymond shows you how to design and create a custom animated pet for Codex. Your pet could be an animal, a miniature virtual version of you, a mascot, or an original character you invent from scratch. You can begin with a written idea, existing character art, a generated image, a brand cue, or another visual reference.

The design process starts with the choices that make the character recognizably yours. You will decide its shape, face, colors, materials, markings, style, and any important props. You will then select one main look as the visual source of truth for every animation.

You will learn how to:

Turn a description or reference image into a clear character design

Choose visual details that remain readable when the pet is small

Create one approved main look that anchors every later pose

Translate the character into animations without losing its identity

Review the motion and correct poses that drift or do not read clearly

Package the finished design as a custom Codex pet

Once the main look is established, you will see it developed into Codex animations for resting, moving, waving, jumping, waiting for input, showing active task work, reviewing completed output, reacting to a failed task, and looking in 16 directions.

The pet is a visual companion rather than another AI assistant. It does not do the work for you, but it gives the Codex workspace a little more personality and makes it feel more like your own.

This session is for people who use Codex and want their workspace to feel less anonymous and more like their own.

Practical AI at Work is a live Bill Talks AI learning series from Cambermast.

Spend one hour, get one vauable result.