AI Use Case and Pilot Management is a project-based course in finding, selecting and running AI pilots with CRISP-DM, stage-gates, the AI Use Case Canvas and business cases, with clear go, no-go or pivot criteria. It suits AI champions, digital transformation teams, project leads and managers who approve AI work, and you leave with a 90-day plan for your pilot, with KPIs to present to leadership.
Many organizations start AI projects with real enthusiasm, yet a good number get stuck in pilots that never finish, known as pilot purgatory, because the problem was never clearly defined, the data was not there to support it, nobody owned the project, success metrics were not set up front, or there was no plan to scale after the pilot. The result is lost budget and lost executive confidence in AI at the same time. This course uses project-based learning. Learners work in groups on a problem from their own work or a case study prepared by the instructor, following the path of a
real AI project. Day one narrows a large pool of ideas down to the most valuable use case using CRISP-DM, stage-gates, the AI Use Case Canvas, a value vs feasibility matrix and a business case. Day two turns that use case into a pilot with a clear scope, KPIs, success criteria, a risk plan, Agile team roles and go / no-go / pivot decision criteria, and closes with a pitch for approval to a review panel. By the end, learners have a pilot charter and a 90-day plan ready to start. (2 days, 6 hours per day, 12 hours in total, Intermediate level.)
Course Details
A 2-day course, 6 hours per day (12 hours in total, 09:00-16:00), run as project-based learning with groups working on a problem from their own work or a case study prepared by the instructor. Intermediate level. No AI technical or programming skills are required. Learners take home a full template set from the AI Use Case Canvas, prioritization matrix and business case through to a pilot charter and a 90-day plan ready to start. For in-house delivery, an optional 1-day Demo Day can be added 4-6 weeks after the training to review pilot results against KPIs.
Section 1: Why Many AI Projects Fail to Scale
Pilot purgatory: pilots that never finish and never scale
Main causes: unclear problems, unready data and no project owner
Often overlooked causes: no KPIs from the start and no change management
Lessons from leading organizations that have scaled AI projects successfully
The impact on budget and on executive confidence in AI
Section 2: The AI Project Lifecycle and Stage-gates
The stages of an AI project following CRISP-DM
Adapting CRISP-DM to the business context and the work of each department
Stage-gates from idea to pilot through to scale
What must be proven before passing each gate
Integrating the AI project lifecycle with the PDCA cycle and a continuous improvement culture
Section 3: Use Case Discovery Techniques
Pain Point Mapping: tracing process pain points from the user perspective
Process Walk: walking the actual process to see frontline problems
The 5 AI opportunity patterns: Predict, Detect, Generate, Optimize and Automate
Building on an AI Use Case Canvas created in an earlier course
Section 4: AI Use Case Canvas
Defining the problem and the users clearly before thinking about technology
Identifying the data required and the right AI approach
Identifying the expected value and the risks of the use case
Setting the metrics that prove the use case has succeeded
Section 5: Prioritizing with Value vs Feasibility
Placing use cases on a value vs feasibility matrix
A scoring model with transparent, measurable criteria
Criteria for business value and data readiness
Criteria for technical feasibility, risk and time to results
Section 6: Business Case, ROI and Build, Buy or Partner
Estimating costs and benefits, both financial and non-financial
Measuring the baseline before starting so results can be compared
Identifying the assumptions the pilot must prove
Criteria for choosing to build, buy or partner
Working with external service providers
Section 7: Day 1 Workshop: From Ideas to Top Use Case
A problem from your own work or a case study prepared by the instructor, worked in groups
Brainstorming at least 10 AI ideas in total
Selecting the best 3 ideas and writing each as an AI Use Case Canvas
Scoring with the prioritization matrix and selecting 1 use case per group
Workshop: prepare an initial business case for the selected use case to carry into Day 2
Section 8: Designing the Pilot
Setting a scope small enough to prove results within limited time
Writing hypotheses and success criteria that can be verified
Choosing KPIs linked to the baseline and the business case
Planning a 6-12 week pilot timeline and the resources required
Section 9: Data and Technical Readiness
Assessing data readiness and planning data preparation
Building an MVP to test the concept with users quickly
Designing for human review of results (human-in-the-loop)
A data and technical readiness checklist before the pilot starts
Section 10: AI Risk and Governance
An AI risk assessment checklist before the project starts
Approval steps and who must take part in the decision
Recording risks and controls in the pilot charter
Linking to the guidance in the AI Governance and Responsible AI for Organizations course
Section 11: Managing AI Projects with Agile
Working in sprints and adjusting the plan from what each round reveals
The roles of the sponsor and the product owner in an AI project
The roles of the data owner, AI champion and IT
Communicating progress to executives and stakeholders
Section 12: Measuring Results and Deciding to Scale
Go / no-go / pivot criteria at the end of the pilot
Planning the scale-up and the handover into production use
Basic MLOps concepts for maintaining AI systems in production
Managing change with users
Section 13: Pitching for Executive Approval
Structuring the pitch as a one-page summary
Data storytelling that shows executives the value and enables a decision
Using Copilot to help prepare documents and slides
Preparing answers on budget, risk and time to results
Section 14: Day 2 Workshop: Pilot Charter & Pitch
Building on the use case and business case selected on Day 1
Drafting the pilot charter: scope, hypotheses, KPIs, team and risks
Planning 90 days of execution with go / no-go / pivot decision points
Workshop: pitch to an in-class review panel for feedback and a quality review of the pilot charter
No public rounds are open right now. Join the waiting list and we will contact you first when the next round opens, or ask us on LINE. Or call 02-570-8449 or 088-807-9770
CRISP-DM has six phases: business understanding, data understanding, data preparation, modelling, evaluation and deployment. The phases are iterative rather than strictly linear; if the evaluation does not meet the business goal, for example, the team goes back to rethink the problem or the data.
Start by writing each use case on an AI Use Case Canvas so the problem, users, required data and measurable outcome are clear. Then score them all against the same criteria, such as business value, data readiness, technical difficulty and risk, and place them on a prioritisation matrix of value against feasibility. Choose a high-value use case that is realistic to deliver as your first pilot.
Pilot purgatory is when an AI pilot keeps running but never finishes and is never scaled up. The usual causes are an unclear problem, data that is not ready, no project owner and no KPIs set at the start. You avoid it by agreeing success criteria and an owner before you begin, using stage-gates to check what must be proven at each point, and making a go, no-go or pivot decision based on actual results.
It suits AI champions and digital transformation working groups, project leads, engineers and section heads who will own a use case, managers who approve AI projects and IT teams who assess technical feasibility. You should have completed AI Literacy for Everyone or understand basic AI concepts and know the frontline problems in your own department. Experience with Kaizen, QCC or Six Sigma is a strong advantage, and no programming is required.
You will be able to find and select use cases with the AI Use Case Canvas and measurable criteria, build a business case with ROI and a build, buy or partner decision, design a pilot with KPIs and a 90-day plan, and present it for approval with a one-page summary, using templates from the canvas through to the pilot charter. Participants who complete the training receive a Certificate of Completion with its own number, which can be verified online.
Yes. It can be run in-house, with group work based on your team's own problems instead of a prepared case study, and an optional Demo Day about four to six weeks later to review pilot results against KPIs. You can request a quotation from this page, or contact us on LINE @itgenius or by phone on 02-570-8449.
AI is becoming a basic organizational capability, just like the internet, cloud and digital transformation before it. Executives do not need to know how to build an AI model or write code, but they should understand what AI is, what it can and cannot do, how…
Factories and manufacturing work are full of documents, data and routine tasks: creating and improving SOPs and work instructions, recording and analyzing production data, finding the causes of defects and downtime, preparing shift and monthly reports,…
Manufacturers facing seasonal demand have to plan under several layers of uncertainty at once: sales that rise and fall through the year, raw material prices, delivery lead times and supplier risk. Planning on experience and spreadsheets alone risks both…
