In total, it’s already been 2.5 years, but for the last 2 years, the fire has really been burning.Every day, I pick up another small piece of knowledge. For me, it’s the 1% method:getting just a little bit better every day. And honestly: I’m proud of myself for that. Sometimes we should allow ourselves to celebrate our own consistency.Because I believe that showing up every single day for 2 years is something not many people on Skool have done. I can only encourage everyone:Stay curious. Keep learning. Keep exchanging ideas. At some point, it will pay off.Maybe financially. Maybe professionally.Or maybe simply because it feels good to understand more about this world. But knowledge is only one side of it. The network is the other. Over the past years, I’ve met so many great people.People to learn with, laugh with, discuss ideas with, or even share a beer with. I’m truly grateful to be part of such a large international network full of inspiring people. Thank you. Holger And now: back to learning.

Following my challenge of semantically grouping 1000+ records I now have a follow-up challenge where I need to cluster a sub-set of the semantic relevance output. As before my current scenario appears limited to processing a few hundred records. (In the video, I show how it processed 238 just 7 days ago and made me think it's done and dusted). But I just tried using the same scenario to process 1300 records and the agent didn't even try; it processed 0 records (effectively giving up before it even began). So this time I want to be proactive and not get myself in the mess that caused me to consume 17000 credits in less than an hour! Thanks to @Aashiya Bi for helping me come up with a 2-part scenario I think will work for large-scale semantic relevance processing. But how do I do something similar for this follow-up scenario. What do I need to change to my current scenario or LLM so that the scenario can execute. Any ideas up front will be most welcome. cc. @Mohamed Jahar

OMG! Did I just get burned. Easily my biggest ouch in the 18 months I've been using Make. I apologise in advance for the poor quality video, (a lag between my voice and screen) but hope you can hear my frustration and start to see my problem. In a nutshell, how do you get a scenario to process batches (of Airtable data) and then end cleanly when it's iterated through all the records. Alternatively, what AI Agent or AI model/LLM could I use (via OpenRouter) that would process 1000s of records). That would be the easiest solution.

Just noticed that Make has added two very useful array functions: ✅ arrayIntersect()✅ arrayDifference() At first glance they seem simple, but they can eliminate a lot of routers, iterators, filters, and custom logic that many of us have been building manually. Example Use Case: Order Fulfillment Imagine you receive two JSON datasets: Orders [ {"ProductNumber":"P1001"}, {"ProductNumber":"P1002"}, {"ProductNumber":"P1003"}, {"ProductNumber":"P1004"} ] Warehouse Inventory [ {"ProductNumber":"P1003"}, {"ProductNumber":"P1004"}, {"ProductNumber":"P2001"}, {"ProductNumber":"P2002"} ] If you want to get the product number that is in both in Orders and Warehouse Inventory you could use ✅ arrayIntersect() OUTPUT : P1003 P1004 Or if you want the product number that is in the first Array you choose (Eg : Orders) and not in the second array (Eg : Warehouse Inventory) you could use ✅ arrayDifference() P1001 P1002 Why This Is Useful Some quick automation ideas: 🔹 Compare ERP products vs Warehouse stock 🔹 Find customers present in CRM but missing in Mailchimp 🔹 Compare Leads from two systems 🔹 Match employee records between HR and Payroll systems 🔹 Find invoices not yet paid 🔹 Compare Shopify orders against fulfillment records 🔹 Detect missing records during data migrations Previously, I would typically use iterators, filters, aggregators, or custom JavaScript for these scenarios. Now it can be done with a single function. Small update, but definitely one of those features that will simplify a lot of real-world automation workflows. Looking forward for your thoughts on this !

The students who succeed aren’t the ones asking: “What’s the best AI tool?” They’re asking: “What’s a problem I can solve with AI today?” Tools change every month. Problem-solving stays valuable forever. What’s one problem you’d like AI to help you solve right now?