How I Diagnose Problems
I don't start with solutions. I start with questions.
- What's the actual business problem? (Not the training request. The problem.)
- What's been tried before? Why didn't it work?
- Who are the real stakeholders? Who has veto power?
- What does success look like in numbers, not feelings?
Most training fails because someone skipped this part.
How I Decide When AI Is Appropriate
AI isn't always the answer. I'm pragmatic about when it makes sense.
I use AI when
- The task is repetitive with real volume
- Quality can be verified before it ships
- The team can maintain it without me
- It actually saves time vs. doing it manually
I don't use AI when
- It's a novelty play with no real ROI
- Stakeholders won't understand or trust it
- It creates dependencies nobody can support
- Manual is actually faster for single tasks
How I Measure Success
Before anything gets built, I define three things:
1
The baseline. Where are we now?
2
The target metric. What should change?
3
The timeline. When do we check?
If we can't measure it, we don't build it. Completions aren't success. Behavior change is.
What I Don't Do
- Hype. I don't chase trends or sell vaporware.
- Vanity metrics. "100% completion rate" means nothing if performance didn't change.
- Drop and disappear. I build things that teams can actually maintain.
- Overcomplicate. Simple systems that work beat complex systems that don't.