Every time our managers propose an AI initiative, it turns into a six-month IT project that goes nowhere. How do we frame and scope these as lean operations-improvement projects rather than machine learning experiments?

Question

Tyler Smith · Accepted Answer

Stop using the words machine learning and artificial intelligence in your project charters. When a project is labeled an AI project, team members treat it as research and development. Timelines expand, software vendors sell custom science projects, and accountability dissolves.

Instead, frame every initiative strictly as an operations-improvement project that happens to use off-the-shelf software. If a department head brings an idea to the leadership team, force them to present it through the lens of a quarterly Rock or a clear operational issue:

- What specific bottleneck or friction point does this solve?
- Which seat on the Accountability Chart will gain capacity?
- How many hours of payroll will this eliminate or reallocate each week?
- What is the measurable metric on our Scorecard that must move within ninety days?

Scope the project to a two-week proof of concept using tools you already own or standard no-code platforms. If a manager cannot demonstrate a working prototype that automates a painful manual step within fourteen days, kill the project. Never build custom models when existing application programming interfaces can solve the problem for twenty dollars a month.

By treating the initiative as a lean workflow fix, you strip away the technology theater. You hold your leaders accountable for business outcomes, measurable labor savings, and streamlined handoffs rather than software exploration. The technology is merely a footnote in service of a more profitable, system-dependent operation.