Browse Dr. Joe Shepherd's writing on AI governance, PE operations, board advisory, and product leadership.
Browse the archive in order, move by topic, or jump to search when you already know the operating question you want answered.
Showing 10 of 22 articles · Page 2 of 2
Most bloated backlogs are really avoidance devices. The three sprint rule forces teams to keep only a short window of real commitment, scope MVPs to fit that window, and let fresh customer signal reset the priorities.
Read →The wrong way to assign AI leadership is by defaulting everything to data science or everything to engineering. The right way is to ask whether the team is still exploring the path forward or already implementing against a clear one.
Read →Most AI board reports describe enthusiasm and pilots. This template forces management to show how AI labor is changing operating metrics, who owns the outcomes, and what decisions the board now needs to make.
Read →When Microsoft licensing shifted to monthly SaaS, our revenue line fell off a cliff. We rebuilt a $1.2B business by redesigning how we attached services to software. The same structural break is happening again as AI compresses SaaS economics.
Read →The smartest AI diligence questions aren't about the model. They're about whether the portfolio company can absorb AI as labor, govern it, and tie it to accountable operating results on the clock that matters.
Read →If AI is delegated labor, the org chart has to show where that labor sits, who supervises it, and how managers are held accountable for blended output. Most companies aren't even drawing that picture yet.
Read →The trust tier model lets AI governance scale with real-world consequence, so low-risk digital labor moves fast while high-risk systems get the control they deserve.
Read →The documentation and controls required for AI diligence are built over time. If AI is part of your value story, starting when the banker calls is already late.
Read →Non-linear careers look messy from the inside and obvious in hindsight. The operating advantage is the pattern recognition they create across systems, not the novelty of the path itself.
Read →AI is widening the gap between organizations that can redesign work, govern decisions, and build usable data foundations, and those still treating AI as a side project. These five imperatives define where leaders need to act.
Read →