The Business Analyst's Guide to AI-Assisted Tool Development

#businessanalysis #ai #mcp

I built a few production tools in 3 months without being a developer. Total time saved per week: 8 hours. Here's exactly how I did it.

Three months ago I was doing this manually:
4 hours/week: Accessibility compliance checking
2 hours/week: Care provider research across multiple regulators
1.5 hours/week: Preprod environment testing
30 mins/week: Rewriting documentation for different audiences

Now I have tools that do it:
🐕 The Moby Accessibility Checker → 30 seconds
🏥 Care Provider Finder → 2 minutes
🔍 MGC Web Page Analyzer → 5 minutes
✍️ Prompt Refiner → Instant

Time saved: ~8 hours per week

I'm 57. Business Analyst . Not a developer.
But I learned how to work effectively with AI to build tools I actually use.
What I built:

The Moby Accessibility Checker
Care Provider Finder
Web Page Analyzer v2.0
Prompt Refiner

All on GitHub. All working. All used daily.

How this happened: Not overnight. Not magic. Proper work:
15-25 hours per tool
Constant testing against real workflows
Multiple rewrites when things didn't work
Real problems solved, not theoretical exercises

What made it possible:
I brought domain expertise. AI brought code generation.
I knew what needed to be built. AI knew how to build it.
I tested everything. AI iterated until it worked.
That partnership is the key.

What I learned:
✅ Start with one painful, repetitive problem
✅ Simple working tools beat complex broken ones
✅ Your BA skills translate perfectly (requirements, testing, iteration)
✅ Documentation matters more than elegant code
✅ Domain knowledge is more valuable than coding skills
❌ Token limits will surprise you (had to refactor entire approach)
❌ Not all data is available (still frustrated about Wales)
❌ First version is never right
❌ Screenshots don't scale (learned that the hard way)

I wrote it all down.
The process. The decisions. The mistakes. The code.
"The Business Analyst's Guide to AI-Assisted Tool Development". Real examples with working code. Real problems and how I solved them. Real failures and how I recovered.

No theory. No hype. Just the practical process that worked.
£5. Launches this week.

hashtag#update Get it here: https://lnkd.in/efVeXn6F

If you're spending hours on repetitive work and wondering "could I build something to fix this?" - this shows you exactly how.

Questions welcome. Still learning myself.