Mind Your Own AI Business

Why this course matters

Most businesses do not need another AI demo. They need an assistant that understands the work, a tool that solves a recurring problem, and a workflow that does not collapse when someone forgets a step.

This course moves beyond casual chatting and teaches professionals how models, context, tokens, memory, instructions, reference files, reusable skills, and tool access work together.

Students use those concepts to build a tested assistant, a reusable SKILL.md, a practical business tool, and a workflow with clear human review.

What you will be able to do

• Explain models, context windows, tokens, memory, instructions, tools, and their practical limits in plain language.
• Map the business knowledge, files, tasks, and decisions an assistant needs to support.
• Create a tested AI assistant with a defined role, instructions, boundaries, reference files, and memory rules.
• Create reusable assistant capabilities using structured instructions and a SKILL.md file.
• Build a small AI-assisted business tool such as an intake helper, estimator, calculator, content tool, or decision aid.
• Design a repeatable workflow with inputs, outputs, tool use, review points, exceptions, ownership, and scheduling.
• Protect confidential information and test assistants and tools for accuracy, privacy, bias, failure cases, and unsafe behavior.
• Measure usefulness and create a practical 30-day plan for responsible adoption.

Seven weeks, session by session

Each weekly topic is taught across two connected one-hour meetings — 14 live online sessions and 14 instructional hours in total. Every meeting combines a short demonstration, guided practice, and one concrete result you keep.

The practice plan

Two classes build each weekly topic. Students adapt every exercise to their own career, business, community project, or creative interest.

Critical thinking skills

• Choose the right AI capability for the task rather than forcing every problem into a chat.
• Decide what information belongs in instructions, reference files, temporary context, or long-term memory.
• Estimate how context and token use affect cost, speed, continuity, and output quality.
• Design boundaries and approval points that keep consequential decisions human.
• Test assistants and tools with normal, incomplete, adversarial, and unsafe requests.
• Evaluate whether a system returns meaningful time without creating equal or greater review work.

How progress is evaluated

This is a not-for-credit course. Learning is evaluated through completion, demonstration, and instructor review of practical work rather than tests or grades. Evidence of learning includes:

• A plain-language map of model, context, token, memory, and privacy choices.
• A business knowledge map and safe reference library plan.
• A functioning, tested business AI assistant.
• A reusable SKILL.md instruction file and capability test.
• A working AI-assisted business tool or prototype.
• A documented workflow with review, exception, and scheduling rules.
• A privacy and testing checklist, business AI playbook, and 30-day implementation plan.

What you need

No textbook is required. Instructor-created system maps, assistant templates, memory guides, token-planning examples, SKILL.md templates, tool briefs, workflow maps, privacy checklists, testing guides, and implementation materials are provided digitally.

Students need a laptop and internet access for the live online sessions.

A paid AI platform is preferred for better results, higher usage limits, memory and file features, assistant building, and tool access. Feature availability varies by provider.

Students must remove confidential and personally identifiable information from all classroom and AI exercises.

Small-business AI operating principles

Where this sits in the catalogue