🧠 Using AI as a Second Brain When Processes Make No Sense at First

Some business processes look simple—until you try to document them.

Then you realise no one does it the same way, the steps are hidden, and the handoffs are undocumented.

I hit this exact wall working on a digital transformation project for a national services provider.

They were replacing three legacy systems with a unified platform.

But the existing process maps were either outdated, inconsistent, or didn’t exist at all.

The team was overwhelmed and stakeholders couldn’t agree on the “real” process.

So I brought in AI—not to replace my work, but to support it.

Here’s how I used it to move faster and get clarity where there was none.

---

🛠️ Step 1 – Turn Messy Notes into Clear Process Drafts

After a few SME workshops, I had multiple pages of conflicting notes.

Instead of trying to stitch it together manually, I used an AI summarisation tool to pull out patterns:

• Frequent steps repeated by different users
• Common terms that hinted at informal workflows
• Inconsistencies in who was doing what

This gave me a starting point.

It was still rough. But now I had a structure I could test with the team.

---

🔍 Step 2 – Use AI to Spot Gaps in Process Logic

Once I built the first version of the process map, I ran each step through a prompt in ChatGPT:

“Given this process, what could go wrong at this step?”

It flagged:

• Missing approvals
• No backup for key roles
• Lack of input validation

I took those ideas into the next workshop.

SMEs confirmed many of them—and added even more edge cases.

That became part of my future-state process and risk register.

---

📝 Step 3 – Draft Requirements from Process Steps

Instead of writing requirements from scratch, I used my process map to feed AI with prompts like:

“Generate a functional requirement for a self-service appointment booking step where the user selects a date and confirms via email.”

It gave me clean, structured outputs.

Of course, I edited them.

But it cut my writing time in half and helped me maintain consistency.

---

🧩 Where AI Helped the Most

• Speeding up early analysis when I had too much info
• Generating ideas for what could go wrong
• Structuring initial drafts of business rules and functional requirements
• Preparing workshop questions based on weak areas in the process

AI didn’t replace my job.

It helped me think faster.

---

⚙️ Tools I Used in the Field

• ChatGPT for summarising notes and generating prompts
• Word Online with Copilot for formatting drafts
• Miro for mapping and testing flows
• Excel to build process comparison tables

---

📈 The Result

The digital transformation project stayed on track.

We built a solid to-be process with stakeholder buy-in.

The requirements document was delivered ahead of schedule.

And most importantly—
The team felt confident in what was being built.

---

🧠 What I Learned

AI is a powerful tool for business analysts, not a threat to us.

It won’t replace the hard parts of the job:

• Talking to people
• Spotting political blockers
• Getting real-world clarity

But it will help you move through the fog faster.

In complex transformations, that edge matters.

---

✅ Final Thoughts

I still do the hard thinking.
I still own the outcomes.

But now I have a digital assistant who helps me connect the dots faster.

If you’re a business analyst or process analyst working in digital transformation, start experimenting.

The results might surprise you.