Execs are drafting posts with it. Teams are running emails through it. Someone added a chatbot to the website and checked the box labeled ‘Add AI’.
But in a lot of cases, it’s just become another step:
Write → Prompt → Fix → Repeat.
That’s not efficiency. That’s extra work dressed up as progress. AI is useful—but only when it’s pointed at the right problems.
AI is an amazing technology and it’s really good at things like:
That’s where I focus. Not replacing people. Not generating noise. But on building systems that take work your team is already doing and reducing friction.
Distillation to Prevent Data Flooding
Most companies record support calls. Almost none of them use that data. There’s just too much of it to deal with so it sits in storage until there’s a problem. Managers might listen (at 3x speed) hoping to catch patterns or rogue reps. It’s simply not sustainable.
A recent project makes every call into usable data.
Managers don’t have to guess what’s happening anymore. They can see patterns across every interaction—not just a handful of reviewed calls. Because the data is onsite, they can still drill down when needed.
Need data distillation? Let’s talk.
Clarifying user intent before presenting results lowers customer service load
Search is another place where AI gets misused. Most systems try to answer immediately—whether they understand the question or not. That’s how you get confident, wrong answers.
A SaaS company brought me in to develop a search engine with a different approach:
It’s a small shift, but it changes the outcome. Instead of guessing, the system helps users get to the right answer faster by using the knowledgebase you already have.
The common thread is simple:
Use AI where it works best.
Keep control where it matters.
Ask yourself: “Where are we spending time on work that could be structured, summarized, or automated?” That’s where you’ll find gains right away.