In the world of business IT we get attracted to the shiny new toy. Right now, that toy is artificial intelligence. Boardrooms are buzzing with buzzwords like LLM, agentic workflows and generative reasoning. The officers are asking anxiously, “What is our AI strategy?”
But here is the bitter truth:
There is no such thing as an AI strategy.
There is only Business Process Optimization (BPO).
“Magic Wand” Illusion
Many enterprises treat AI like a magic wand. They believe that by implementing a sophisticated neural network their structural inefficiencies will disappear. They think AI brings intelligence.
It’s not like that.
Like every major technological change before it – from the Steam Engine to spreadsheets – AI doesn’t inherently make an organization smarter. AI, like any other tool, only gets faster.
If you automate a stupid decision, you make stupid decisions at lightning speed. If you apply an agentic AI workflow to the bureaucratic nightmare of an approval chain, you haven’t fixed bureaucracy; You’ve just created a robot that hates its job as much as your employees.
unstructured data trap
However, AI has one superpower that previous tools lacked: This is the first technique that is really useful for handling unstructured data.
For decades, traditional software demanded structure. Rows, columns, booleans and fixed fields. If the data doesn’t fit in the box, the computer can’t read it.
AI changes this. It can read messy emails, interpret ambiguous Slack messages, parse PDFs, and analyze images. But this capability highlights a huge, hidden problem in most enterprises.
Processes that rely on unstructured data are generally unstructured processes.
Because computers couldn’t handle chaos, humans took over (before AI). And humans don’t always follow flow charts. These processes – such as “handling a complex customer complaint” or “brainstorming a marketing campaign” – are often ad-hoc, spontaneous, and completely undocumented. They live in the minds of your senior employees, not in your SOPs.
You can’t automate what you haven’t designed
This brings us back to BPO. You can’t apply AI to these “hidden” processes unless you bring them to light.
If you want to use AI to process unstructured data, you must first bring structure to the workflow. You need to improve your process design to take into account the ambiguity that AI can handle.
ask yourself:
- What is a trigger? (Where does unstructured dirt come from?)
- What is change? (What exactly is a human—or now AI—supposed to make or conclude from that mess?)
- What is structured output? (How does this flow back into your hard ERP or CRM system?)
speed vs intelligence
Let’s clarify the difference between “clever” and “fast.”
Intelligence implies knowledge, context and subtlety. While AI models are simulating logic better every day, in a business context, they are fundamentally pattern-matching engines. They excel in acceleration.
- Old way: An analyst reads 50 contracts (unstructured), highlights risks based on gut feeling (unstructured process), and summarizes them over 3 days.
- AI Method: An AI scans 50 contracts and extracts specific risk segments based on defined parameters in 3 minutes.
Process (Review contracts -> Identify risks -> Summarize) hasn’t changed, but had to be rigorously defined for AI to work. intelligence (Knowing what “risk” actually means) still requires human governance. what has changed is velocity,
bottom line
Stop chasing hype. Stop looking for any particular “AI savior”.
Go back to the whiteboard. Map out your value chain—especially the messy, human-centric parts involving unstructured data you previously ignored. Find the obstacles. Identify waste.
Once you have a well-organized, logical, and robust business process, Then Apply AI to hit the accelerator.
Technology changes.
Rules of commercial efficiency do not apply.
It’s always the process, idiot!
And that’s where real AI tools are missing the point, because they weren’t built for that.
The Great IT-Divide: Why AI-Adoption in Enterprises is Failing
IT innovation moved from business tools to social technology, creating two distinct IT worlds: Business-IT (compliance, efficiency) and Social-IT (social interaction). Understanding this divide is important for enterprise adoption and explains why businesses struggle with adopting AI.
Live long and be prosperous 😉🖖
<a href