- AI Unhyped
- Posts
- How to find hidden opportunities for AI
How to find hidden opportunities for AI
A real-world look at mapping processes, and finding ROI in unexpected places.
Dino Gane-Palmer
March 18, 2025 • Estimated Reading Time: 9 minutes
Hi, and happy Tuesday.
They say healthcare and trucking are slow to adopt new tech, yet two leaders in these “old school” industries have discovered an approach to technology adoption that saves millions.
💡 By the end of this post, you’ll see exactly how they do it—in a way that avoids expensive failures, wins employee buy-in, and identifies the right areas for automation before writing a single check.
🔥Skip to the end if you just want to steal this week’s AI prompt.
✉️ Reply to this mail with your questions about this week’s content.
IN THE SPOTLIGHT
Picture a busy medical practice—patients flipping through magazines, phones ringing off the hook, staff hurriedly shuffling paper forms. Across town, in a warehouse full of truck parts, you hear the hum of machinery and see workers lifting heavy gearboxes, scanning shelves, and tracking orders with scribbled notes on clipboards.
It might seem these two worlds have little in common, yet both face the same challenge: identifying where technology (including AI) truly helps, and where it’s just hype.
In a recent interview, I spoke with two industry leaders making waves:
Dan Hosler: Co-founder at DuneGlass Capital and CEO of Allied OMS, steering over 48 oral surgery practices in 12 states.
Michael Chudacoff: Former President of Midwest Trucking & Auto Parts, which was recently sold to private equity.
They’ve shared a powerful lesson: “Map processes first, then identify where AI or new tech can help.”
Why Traditional Companies Struggle with AI
Big talk about AI often conjures images of self-driving trucks and robotic surgeons. Those breakthroughs are flashy, but in the day-to-day reality of healthcare clinics and distribution warehouses, many initiatives flop.
We often see a “solution in search of a problem” scenario—leaders buy advanced tech only to discover staff haven’t mapped their everyday tasks or data flows. The result? Underutilized software, staff confusion, and wasted spending.
Imagine you’re trying to build a new house, yet you skip the architectural blueprint entirely. That’s what happens when companies chase AI without first documenting how processes actually work.
“We had to start with mapping out the processes—everything from consult to surgery to post-op. Only then could we see where AI or automation helps. In healthcare, if you jump to automation without that understanding, you end up with misaligned solutions.”
“I had departments with 10, 20, 30 people. Sure, AI can speed things up, but if you don’t map out tasks first, you’re just dropping complicated tools on top of old workflows—and that never sticks.”
For both leaders, the biggest hidden barrier wasn’t the tech—it was the messy tangle of everyday procedures nobody had written down. And that’s precisely why so many big AI projects go nowhere.
Still, both caution that the right approach—meticulous process mapping—demands time and patience, which conflicts with the hype that “AI will fix everything, right now.”
The “Map First” Method in Action
The breakthrough insight? Before even talking AI, get clarity on who does what, how, and why.
Document the Flow
Both Dan and Michael mentioned physically or digitally mapping the steps in patient scheduling, insurance verification, picking parts in a warehouse, or answering the phone after hours. Only after you see each handoff and repetitive task can you spot precisely where AI (or any tech) can reduce bottlenecks.Identify High-Leverage Tasks
Dan discovered that mundane tasks like insurance validation were time sinks. AI chatbots or RPA (robotic process automation) could handle eligibility checks, letting staff focus on patients.
Michael saw repetitive quoting processes for trucking clients. He realized an AI-driven tool could quickly generate quotes from historical data, but only if the data was properly captured first.
Check for Employee Readiness
Fear of “being replaced” can torpedo adoption. By framing AI as an augmentation (not a replacement), you get buy-in.
“Older staff can be skeptical. I had to show how the tech would make their jobs easier, not just cut headcount.”
Test Small, Then Expand
Dan runs pilot programs in 5-7 practices before rolling out changes system-wide. Similarly, Michael introduced new ERP modules in a single location. Once proven, they scaled it.
It’s easy to dream about flashy robotic arms in surgery or fully automated truck loading, but you’ll save time, frustration, and money by following these four steps first.
Thinking of your own processes? Reply to this email if you want to walk through the approach step by step.
Action Steps for Leaders
Step 1: Map Your Current Workflows
Grab sticky notes, a whiteboard, or a digital tool. Diagram each action, handoff, and decision point.
Step 2: Ask “Where’s the Bottleneck?”
Look for tasks that take forever, get repeated hundreds of times, or cause the biggest backlog.
Step 3: Evaluate Small AI or Automation Fixes
From chatbots for data entry to scanning software for inventory, test minimal products first. No million-dollar investments needed.
Step 4: Monitor & Measure
Did the pilot free up staff time or cut errors? If yes, scale up. If not, pivot with minimal sunk cost.
Want help with your process mapping?
Try this prompt with your favorite AI chatbot:
I am in the <your industry> industry working in <your department>. I want to create a detailed process map of our current workflows, from start to finish, for <specific process>. My goal is to identify the parts where there is recurring, routine intellectual work that could be accelerated or automated with LLMs. Please ask your 3 most important clarifying questions then output the process map using ASCII code. Highlight the parts where there may be greatest potential based on intensity of intellectual time/effort.
Here’s what I got when I replaced <your industry> with “manufacturing”, <your department> with “R&D” and <specific process> with “technology vetting”.
Because technology moves faster than we can plan, the truly risky move is to invest blindly. By mapping your processes first, you shine a light on the hidden inefficiencies AI can solve.
I’d love to hear your thoughts, especially if you’re already mapping workflows or exploring AI tools in your org. Reply to this email to share your experience or ask any questions.
Best,
Dino
FROM OUR INTERVIEW SERIES
 | 🎯 What you’ll learn:
|
 | 🎯 What you’ll learn:
|
 | 🎯 What you’ll learn:
|