Most conversations about AI in business software start with the technology and work backwards to a use case. The ones that actually ship something useful start with a specific, annoying, manual task and work forwards.
The tasks worth automating first
AI earns its place fastest on tasks that are repetitive, high-volume, and tolerant of the occasional miss with a human in the loop to catch it — drafting, summarising, classifying, and surfacing information someone would otherwise dig for manually.
Where it adds more risk than value
Tasks with a high cost of error and low tolerance for a human check — financial approvals, compliance decisions, anything customer-facing and irreversible — are a much harder bar to clear, and often not worth the added complexity yet.
“The best AI features are invisible: the user just notices the tedious part of their job got shorter.”
A practical filter
Before building an AI feature, ask what the manual version of the task costs today, and whether an imperfect but fast automated version is genuinely better than that baseline. If it isn't, it's not ready yet.
AI is a tool for removing specific, well-understood friction — not a feature to add for its own sake.