Skip to main content
Underpin WorksDiscuss an AI opportunity

AI Strategy

Why we start with the business case before the AI.

Most AI projects don't fail on technology. They fail because the workflow wasn't worth changing in the first place. A practical way to weigh expected value against the cost of change — before anything gets built.

Prateek BawaPublished 14 July 2026Updated 14 July 20268 min read

This article is being written. The outline is set and it will be published under AI Strategy.

Working on something this touches on?

Discuss an AI opportunity

Written by Prateek Bawa, who designs and builds the systems described here. More about Prateek →

Related articles

AI for RecruitmentWhy recruitment agencies struggle to maintain consistent business developmentDelivery pressure, late-recognised opportunities and manual follow-through — where the BD execution gap actually comes from.7 min readAI for RecruitmentCandidate-led business development: turning your database into a commercial assetThe most credible reason to start a client conversation is often already in your candidate network.9 min readAI StrategyHow to decide whether a workflow is worth automatingA practical framework: expected value against cost to change, and why 'easy to automate' is the wrong starting question.6 min read