The main families of technology sold under the AI label, from rules-based automation and predictive models to generative AI and agents, and what each is good for. Then the automation spectrum, from tools that assist a person to systems that act on their own, and where human judgment belongs along it. This is the vocabulary you’ll use to size up your own use cases and every tool you evaluate afterward.

Before looking for use cases, decide what AI is supposed to do for your business. This module ties adoption to your business goals, defines how much risk you’re willing to take on, assesses your company’s capacity to absorb change, and identifies the conditions under which the right answer is to wait. The result is a short strategic frame that tells you which opportunities are worth pursuing and which are just efficient distractions.

How to look at your own operation the way an outside advisor would: tracing where time, money, and errors accumulate, separating the symptom people complain about from the process causing it, and testing each opportunity against the strategic frame you set in Module 1. You come out with a prioritized set of use cases grounded in how your business works.

Every AI use case runs on data, and most stall because the data isn’t where, what, or as clean as everyone assumed. This module covers how to find out where your data lives, whether it’s complete, consistent, and accessible enough to support the use cases you’ve identified, who owns it, and what it would take to close the gaps. You’ll sort your candidates into what’s viable now and what needs groundwork first.

Real AI failures, from chatbots that made promises their companies had to honor to hiring software that screened out qualified applicants, alongside the failures most common in small businesses: tools nobody uses, pilots that never scale, and subscriptions that don’t connect. Each case is traced to the decision that could have prevented it. You finish learning how to conduct a pre-mortem on your own top use case, working out the most likely ways it could fail before you commit to it.

A live session at the course midpoint. We walk through a realistic small business’s strategy, use cases, and data gaps using the frameworks from Modules 1 through 4, then open the floor for questions.

What data can touch which kind of AI tool and how to read a vendor’s data-use terms, plus who’s accountable when an output is wrong, when a human has to review something before it reaches a customer, how to spot inequitable outcomes in an AI-assisted decision, and who owns what the tool generates.

A scoring rubric that separates real capability from a well-run demo, plus the questions a sales call is built to dodge.

Take your strongest candidate from Modules 2 and 3 and scope it as a real pilot: owner, timeline, success criteria, and the conditions under which you’d stop it. Then build a roadmap around it, whether you’re starting fresh or pulling existing tools and experiments into a plan that finally connects them.

How to bring skeptical staff, partners, or investors along: explaining why you’re making the change, addressing concerns about jobs and quality, and training people so what you’ve chosen gets used.

The difference between adoption metrics and outcome metrics, how to read your pilot’s results against the criteria you set in Module 7, and a lightweight way to decide whether to scale, adjust, or shut it down.

A closing live session. The same example from the previous live session is carried through governance, vendor selection, roadmap, team buy-in, and measurement, followed by open Q&A on anything from the full course. Volunteers can bring their own pilot plan or roadmap for discussion.