Someone built an AI planning tool with Claude that flags sudden demand spikes and runs 250 test scenarios. No developers needed. Small businesses can customize it themselves instead of using spreadsheets.
That work has taught me one thing before anything else: planning is not a spreadsheet exercise. The companies running planning on Excel and emails, this is where this is going to help them.
Instead of talking about AI in theory, we invite practitioners who are actually building tools, workflows, and applications with AI to show us what they have built, how they built it, and what it could replace. It's a real practitioner walkthrough of a tool built to solve real planning problems.
So what I built is 3 things: low cost, fast to deploy, and customizable tools. So here is the part that I want to spend a minute on because it matters on how small and medium enterprises should think about this kind of tool going forward. There is no developer in the middle, no translation layer. It's only me and Claude. I'm telling him what I want in plain language, like I'm talking to you right now, and Claude is executing. I have built exception reports in this engine Because when you have so much SKU, so many branches, it's so hard to follow the demand signal. You need to manage by exception. And when I say by exception, for example, if you have an SKU that's selling 10 units per month, suddenly you see a spike that's 40 units per month, this is an exception that's going to get flagged. For example, for this one, I have 250 scenarios tests to validate the data. I created 250 test plans scenario. Like any implementation, you need to go through test plan scenarios and we did the test plan and for that tool that I implemented, he will make mistakes, but he will convince you about those mistakes.
So you will make mistake and tell you that you're basically wrong. This is where you need to stress test.
You need to have good test scenarios. But if you want to save on those and you're a small and medium enterprise where time, cost, and customization is valid for you, is valuable for you, this is your place. You understand your business better than any developer that's going to come and start doing or start architecting the solution for you.
I want to thank you, Alain, again for sharing this, and that was very interesting.
