Running an AI project: from use case to measured pilot

Objectives

• Write a scoping note for an AI project: use case, cost and benefit estimate, outcome indicator, risks.
• Write the scope of an agent or automation: what it does, what it does not do, what is escalated to a human, and which irreversible actions require approval.
• Design a before-and-after measurement protocol on a limited scope, with a volunteer team and a fixed duration.
• Write an end-of-pilot report concluding with stopping or rolling out, backed by data.
• Draw up the acceptance checklist for a deployment: source code, documentation, transferred access, trained teams, a costed maintenance contract.

Audience

Project managers, business owners carrying an AI or automation project, IT directors and IT managers, quality or transformation managers.

Prerequisites

Lead or have led a project in a company, in any field. IA-110 is recommended for executives but not required. Equipment: one laptop per participant; for remote delivery, a stable connection, microphone and camera; a real AI or automation project, under way or planned, described on one page and brought in the morning.

Format

In person, synchronous remote or hybrid; in-company or open enrolment; 6 to 12 participants. Duration: 7 hours, 1 day; remote, 2 sessions of 3 h 30 one week apart, the scoping note being written in between.

Assessment

Beforehand, a placement questionnaire. During the course, a check quiz and a case study (decide on an imposed pilot report). At the end, the scoping note assessed against a grid, an end-of-course evaluation and a named certificate of completion. At three weeks, a follow-up review of project progress.