Automation · Coming soon
Asked to bring AI into operations? Start with the work you repeat
You map your team’s recurring work, score which jobs are worth automating, and build two or three AI workflows. Each one has a named owner, a human check before anything leaves the team, and a before-and-after time measure you can take to leadership. The workflows use no-code tools an operations team can run without an engineer.
The course: AI workflow automation for an operations team
$25 $49 early-bird price
This course is being built now. Join the early-access list to reserve the early-bird price and shape what it covers.
The course is in production. Joining the list is free, and you only pay if you buy at launch, at the early-bird price.
If this sounds like your week
- The same data gets copied from one tool into another by hand every week, and every copy is a chance for an error.
- Leadership has asked what AI will do for operations, and you have no plan you could defend in a meeting.
- People are trying ChatGPT and Zapier on their own, and none of it is shared, written down or owned.
- When an automation someone built breaks, nobody knows how it worked or who should fix it.
What you do now, and where it stops working
Today the team absorbs the work by hand, a spreadsheet is the system of record, and the one person who built an automation is the only one who can fix it. That holds until they go on leave, or until leadership asks what the AI effort has produced and there is no number to give.
What changes after the course
- A one-page map of your team’s recurring jobs, with the hours and the error cost of each.
- A ranked shortlist that shows leadership why you are automating these jobs first and not others.
- Two or three AI workflows running in the team, each with a named owner and a person who approves output before it goes out.
- A before-and-after count of hours per job that you measured, not estimated.
How it works: Map, score, build, own, measure
- 1Map: log every recurring job for two weeks with who does it, how often, how long it takes and what goes wrong.
- 2Score: rate each job on time cost, error cost, how clear its rules are and what a mistake would risk, then pick the top two or three.
- 3Build: start from the intake, triage or reporting template, with AI on the steps that need judgement and plain rules everywhere else.
- 4Own: give each workflow an owner, a review step and a written runbook, so it survives the person who built it.
- 5Measure: time the job before and after over the same period, and report the difference.
What you get
- Work mapping worksheet
- Automation scoring matrix
- Workflow templates for intake, triage and reporting
- Workflow ownership and review template
- Video walkthrough of an operations workflow from map to measurement
This is for you if
- You lead or manage an operations team of three or more people.
- Someone above you has asked what AI will do for operations.
- Your team moves data between tools by hand every week.
Skip it if
- You need an enterprise AI platform rollout with procurement and integration engineering. That is a project, not a course.
- You want AI to replace people. Every workflow here keeps a person approving the output.
- You work alone. The small business course fits you better.
Before you join
- Is this more AI hype I will have to explain away later?
- It commits to a map, a ranked list, two or three workflows and a measurement, nothing larger. MIT NANDA’s 2025 study found 95% of organisations it researched saw no measurable return from generative AI, and tied that mainly to brittle workflows and poor fit with daily work, which is why this course starts from the daily work. Jay West leads DevOps for the marketing organisation of a multi-location specialty retailer and builds AI agent systems with review gates.
- Nobody on my team can build this. Do we need an engineer?
- No. The workflows use no-code tools and ready-made templates, so the builder needs to be comfortable in a spreadsheet, not a code editor. Where a step uses AI, the template includes the prompt and shows you how to test it.
- What happens when a workflow breaks, and who fixes it?
- Each workflow has a named owner, an alert when a run fails, and a runbook that says how to do the job by hand while it is fixed. Nobody has to reverse-engineer someone else’s automation at a bad moment.
- Will IT or security block this?
- The scoring step flags any job that touches customer or employee data. The course gives you the questions to settle with IT before you build: where the data goes, which accounts the workflow runs under, and who can change it. Settling it first keeps the workflow from becoming a tool IT discovers later.
- My team will not change how they work. Why would this be different?
- The course has you pick jobs the team already dislikes, and the person who does the job today becomes its reviewer. The change is in what they check, not whether they are needed. No course can promise adoption; this one is built to make it the easy choice.
Rather have it built for you? Describe the project and the same work can be done for your team.