Build Your Own AI Assistant: A Systematic Way to Elevate the Work You Repeat

Paige Bradbury

CEO, Principal Learning Architect

Stop starting from scratch every time you ask AI for help

Most of us are still asking AI for help one prompt at a time, and that gets old fast. You explain the task, the audience, and the standards, then do it again in the next chat.

An AI assistant gives recurring work a standing brief. But useful assistants need more than a role and a friendly name. They need clear instructions, trusted reference material, quality standards, and rules for when to ask instead of guess.

In two hours, build and test a first version of an AI assistant for one recurring workplace task. Leave with your own reusable system instructions, initial test results, and clear next steps for refinement and responsible use.

You don't need coding skills or a previous TBG course. Neither do you need to know it all.

You stay the decision-maker: the assistant does the repetitive work, and you bring the judgment.

What you’ll learn

Move from repeating prompts to building, testing, and refining an AI assistant for a recurring workplace task.

  • Define its purpose, users, scope, and the work that stays with people.

  • Choose approved reference material and specify how the assistant should use it.

  • Turn your requirements into plain-language instructions, quality criteria, and examples.

  • Specify when the assistant should ask questions, flag missing information, or stop for human review.

  • Test normal, vague, conflicting, and out-of-scope requests against your quality criteria.

  • Revise the instructions, log what changed, and identify what needs checking before workplace use.

Workshop agenda

  • Choose the job and define the boundaries (15 minutes)

    Choose one recurring task. Define its users, a useful result, and what stays with people. Briefly connect a one-off prompt to standing system instructions.

  • Write the assistant's standing instructions (30 minutes)

    Draft its role, context, steps, quality criteria, and examples in plain language. Separate what it does from the standards every response must meet.

  • Add reference material and human review rules (20 minutes)

    Choose approved sources. Define what to do when information is missing or conflicts, and when the assistant must pause for human review.

  • Build and stress-test the first version (30 minutes)

    Build your assistant, then test normal, ambiguous, out-of-scope, and conflicting-information requests. Check the responses against your quality criteria.

  • Repair, retest, and plan the next use (15 minutes)

    Revise a weak instruction, repeat the failed test, and record what changed. Identify remaining limitations and checks needed before workplace use.

  • Questions and next steps (10 minutes)

    Clarify your next test, human review responsibilities, and any workplace approvals needed. Leave with a first version and a practical refinement plan.

Learn directly from Paige

Paige Bradbury

Paige Bradbury

Principal Learning Architect, Trainer, Speaker, Writer

Course Creation & AI Consulting For
Microsoft
Coursera
SAP
National Center for Education in Maternal and Child Health
Xbox
See all products from Paige

Who this workshop is for

  • Professionals who repeat the same work, including AI literacy cohort alumni ready for the next step, who want AI to handle repetitive tasks.

  • Leaders, L&D professionals, and workforce development teams building reusable AI support for others, with human oversight built in.

Prerequisites

  • Basic AI chat experience

    You have used an AI chat tool and can enter a request, review its response, and ask follow-up questions.

  • Access to an AI tool

    Bring your own AI account. Saving an assistant requires builder access; otherwise, draft and test instructions in a fresh chat.

  • One recurring task and safe sample material

    Bring one recurring task. Use fictional or de-identified examples unless your organization permits the material in your chosen AI account.

What's included

Paige Bradbury

Live sessions

Learn directly from Paige Bradbury in a real-time, interactive format.

Certificate of completion

Share your new skills with your employer, on LinkedIn, or on your portfolio.

Two-hour live build workshop

Build and test a first-version assistant for one recurring task. Create your own reusable instructions, review initial results, and identify what to refine before workplace use.

Optional follow-up working session

Join a one-hour follow-up on Saturday, November 28 at 12 pm ET to showcase your assistant, ask questions, or troubleshoot what you’ve tried. Included at no additional cost. Attendance is optional; the core workshop stands on its own.

Maven Guarantee

Your purchase is backed by the Maven Guarantee.

Frequently asked questions

Maven for Teams

Reimbursement

Get your company to pay

Everything L&D needs: email template, receipts, and certificate of completion.

Get reimbursed

Team discount

Learn with your teammates

Save 20%+ when 2 or more teammates enroll in the same cohort.

Save 20%+ with a team

Private cohort

Run a cohort for your org

A dedicated cohort with a custom schedule and curriculum, tailored to your team.

Book a private cohort

$397

USD

5 days left to enroll

Nov 19
·

12–2pm EST

Enroll