Skip to lesson content

Where a business analyst meets AI

14 min

Generative tools are good at producing likely text. Analysis is good when it is true to a source: a meeting, a system, a policy, or a number someone can check. Those are different jobs.

Use the model where wording is the bottleneck. Keep the human where commitment is the bottleneck: scope, priority, and what "done" means. If you cannot point to the source of a sentence in a spec, it is not ready.

Jobs to delegate, jobs to keep

  • Delegate first drafts, outlines, and regrouping of notes you already have.
  • Keep priority, scope cuts, and anything a stakeholder must confirm.
  • Label AI-assisted text until you have checked it.

Watch Google Cloud Tech, "Introduction to Generative AI," from the opening definitions. This lesson maps that idea onto BA work: drafting and clustering, not deciding for the stakeholder.

Three boxes labeled Notes, Draft, and Decision
The model drafts from notes. You still make the decision.

Practice

List three BA tasks from your week. Mark each as draft-with-AI or decide-yourself. If a task mixes both, split it.