We just dropped a new Dial In Podcast episode with AI for Contractors community member Keith Young from Young Electrical Services. Keith's been with us for about three years now. Starting as two guys in a Dodge Ram, now he's running 17 people with a shop, estimators, project managers, the whole thing. We got into a lot in this one, but the part I keep coming back to is how he talked about AI on the job. Keith used to spend four days reading through spec sheets before quoting a project. That was fine when it was just him. But once he's paying estimators and PMs, four days of wages to read paper starts to hit different. Now his team downloads specs into AI, asks a few targeted questions, and gets through it in under an hour. They still double-check everything. Still read the drawings. They just stopped burning days where they didn't need to. He's not running his whole company through AI. He's not trying to automate everything. He just found the spots where he was losing time for no reason and plugged AI into those gaps. If you haven't tried it yet with your own specs or project docs, that's a solid place to start. Upload a PDF, ask it what the scope covers, ask it to flag anything unusual. See what comes back. Full episode is on YouTube. Worth the watch if you've got a team and you're really feeling what their time costs.

Anyone else burning through their Claude Pro tokens way faster than expected? Coming from ChatGPT Plus I never ran into this — curious if others are noticing the same thing or if there are tricks you're using to stretch usage further.

We just added something to the Classroom you're going to want to check out. 👇 If you've been in the last few sessions, you've heard us talk about Skills, the feature that turns AI from a tool you have to explain yourself to every time, into something that already knows the job before you start. We broke it down into a full module inside the Classroom: ✅ What a Skill actually is ✅ Why it matters for your contracting business ✅ How to find one, build one, and load it into Claude in under 5 minutes Contractors in the group are already using Skills to generate RFIs, change orders, project kickoff checklists, and follow-up emails, without re-explaining themselves every single time. If you do a task more than once, it should be a Skill. Head to the Classroom tab and look for the AI Skills module, it's all in there. https://www.skool.com/ai-for-contractors/classroom/ea27606f?md=41ade066166c47f683e7eebd99ada75c Drop a comment below once you've had a chance to go through it. Would love to know what Skill you're going to build first. Stay Dialed.

🎉 COMMUNITY WINS THIS WEEK 🎉 - Shamus shared how he onboarded a demo crew in under an hour using NotebookLM flashcards and a generated audio walkthrough. - Keith uploaded a 212-page addendum and pulled out only the electrical scope, saving his team roughly three hours. - Ryan connected his CRM and Google Calendar to Claude and now runs his full appointment schedule, customer info, and directions through one screen without opening his calendar or CRM separately. NOTE: Unfortunately our meeting recording software crapped out after 35 minutes. The full transcript is attached if you'd like to upload it to claude and ask questions about the second half which heavily covers Claude Skills. Thanks everyone for dialling in at 7PM on a Tuesday. Attached is the recording & takeaways for those who couldn't make it or want to listen back. Summary and list of tools with links below. See you next Tuesday 🔥 Dial in. 📓 NOTEBOOK LM DEMO PROJECT: If you want to follow along or play around with the demo from last night's session, here's the project link - https://notebooklm.google.com/notebook/cc2789c1-bc75-40c3-a733-6ea3d70af222 🛠️ THE AI TOOL BELT: The AI Tool Belt is an early-access app Scelta is currently developing that gives contractors an easy way to find, create, and organize AI skills for their trade business. As part of the community you get access to it while it's in development - ai-toolbelt.vercel.app/discover 🤖 AI SUMMARY: This was a double-down night. Instead of introducing new tools, we went deeper into NotebookLM and Claude, and covered actual use cases from group members. David walked through a full NotebookLM demo using a real Windsor renovation project with 47 uploaded sources. We covered how to ask the AI questions about your project documents, save answers as notes, generate slide decks, flashcards, mind maps, and even a podcast that explains your project while you drive to the job site.