Plain answers before you enroll.
This page explains what The Practical AI Workbench self-paced Builds include, what they do not include, how access and support work, what completion means, and how to start safely with AI tools. Live workshops are scoped separately.
What one Build includes.
Each standalone Build is designed around one finished artifact, so a learner can buy one Build without needing the whole library first.
Class Path
A guided sequence that tells you what to open first, what to build, and when you are done.
Downloadable Pack
Templates, checklists, examples, practice labs, capstone instructions, rubric, and checkpoint questions.
Finished Artifact
A reusable prompt pack, context packet, Work OS starter, surface map, agent pack, or workflow SOP. Workshop materials are part of scoped live delivery, not the default class model.
Which Build should I buy first?
Use the Build that matches the problem you already feel. The full path is useful, but one Build should still solve one real bottleneck.
Start With Prompts
If AI output feels generic, start with Prompt Foundations.
Start With Context
If your files, notes, drafts, or screenshots are messy, start with Make Your Work Legible.
Start With Work OS
If you keep re-explaining your work, start with Build Your AI-Ready Work OS.
Start With Tool Choice
If you are tool-confused, start with AI Surfaces and Harnesses.
Start With Agents
If one recurring task might deserve a helper, start with Basic Agents.
Start With Workflows
If one recurring process needs a repeatable path, start with Agentic Workflows.
Start With Delivery
If you need to teach or facilitate this work, ask about a live workshop or enablement session.
Access, refunds, support, and completion.
Access
Standalone Builds include access to purchased materials and included minor updates for the stated access period.
Refunds
Use a 7-day fit-based refund window for genuine mismatch, duplicate purchase, access problems, or cases where the Build does not match the description.
Support
Self-paced Builds include materials, templates, examples, rubrics, and checkpoints. Included support covers access issues, missing files, broken links, and basic start-path confusion.
Not Included
Unlimited consulting, done-for-you implementation, private file cleanup, custom automation, guaranteed outcomes, professional accreditation, and autonomous agents acting without review are separate from the self-paced Build.
How to tell whether the Build is a fit.
A self-paced Build should give you a useful first move without needing a private walkthrough. Use these signals before buying another class, asking for extra help, or deciding it is not a fit.
Good Fit
You can name the artifact, open the start file, try the first 15-minute action with low-risk material, and see how the rubric will help you judge the result.
Start-Path Support
If a link, file, download, or first step is unclear, ask for access or start-path help before assuming the Build itself is wrong.
Not A Fit
The Build may not be right if you need done-for-you implementation, private artifact review, sensitive-data handling, custom automation, or a tool-specific setup service.
Refund Signal
Use the 7-day fit window for genuine mismatch, duplicate purchase, access problems, or a Build that does not match the description.
Tools, privacy, and safe starting points.
The Builds are platform agnostic. ChatGPT, Claude, Gemini, Codex, Claude Code, Antigravity, MCP, CLI tools, and similar systems are examples, not requirements.
Use Bounded Examples
Start with low-risk examples, folder listings, fictional samples, or redacted material before sharing real private content with any AI tool.
Read-Only First
For file and context work, use AI to propose names, buckets, exclusions, and packets. Review before making real file changes manually.
Keep Sensitive Data Out
Do not upload passwords, financial records, medical records, legal documents, HR files, student records, client secrets, account exports, or regulated material unless you have permission and a safe environment.
Quick FAQ.
Do I need to take the Builds in order?
No. Each $49 self-paced Build is designed to stand alone. Use the full path if you want the complete The Practical AI Workbench artifact chain.
Will the Builds become outdated when tools change?
Some examples may change. The core artifacts should stay useful because they focus on work structure: prompts, context, source maps, decision rules, permissions, tests, SOPs, and review gates.
Is this a certification?
Not in the professional accreditation sense. Completion should be described as a Certificate of Completion or Verified Completion Badge, and only after the learner has built the artifact, reviewed it against the rubric, passed the applied checkpoint, and saved the completion record.
Can I share completion on LinkedIn?
Yes, if the completion record is real and modest. The safest wording is: Completed [Build Name] from The Practical AI Workbench. Built [artifact], reviewed it against the Build rubric, and passed the applied checkpoint. Do not describe it as external accreditation, professional licensure, guaranteed job readiness, or permission to deploy autonomous agents without review.
Can AI rename, move, delete, or change my files?
The class pattern is read-only first. AI can help propose changes, but learners should review the proposal before applying any real file change manually.
What if I get stuck?
If a link, file, download, or first step is unclear, ask for access or start-path help. If you need someone to adapt the artifact to your real work, review private material, configure tools, or build the system with you, that is a separate live help or implementation offer.
Why not just sell a subscription?
A subscription only makes sense when there is living value: clinics, updated examples, tool-change notes, member templates, implementation prompts, and light review. The simpler first model is pay-per-class or a focused bundle.
Before buying, these questions should be answered.
What artifact will I build?
Every Build should name the finished deliverable before access is requested.
What file should I open first?
Every paid delivery should point to a paid-buyer start file or bundle bridge guide.
What support is included?
Access and start-path help are included with self-paced access; custom work is separate.
What is the refund window?
The current standard is a simple 7-day fit-based refund policy.