What happens when you stop asking whether AI can help your coaching business and start asking which parts of your workflow it can simply take over? Afro and Farah open this second instalment mid-experiment. Afro has already used Claude to pull a transcript from a Zoom recording, extract the key points from an hour long client call, and generate a pitch deck, all in a single session. The design was rough, he admits, but the content was 100% accurate. His fix: pipe Claude's output into Canva instead of using Claude Design directly. The bigger story is the podcast workflow they are building right now. Afro walks through the chain: upload the episode to Descript, trigger the agent, it pulls the transcript, drafts the summary and ten quotes, pauses for approval, schedules to Spotify and all major platforms, then surfaces ten short clips edited with preset captions via Opus, pauses again, and schedules those clips to social media. He puts the completion at 80%. The only human step left is pressing record and saying "kick it off." Farah's content pipeline is close behind. She records a short video, drops it into a folder, and the agent adds a thumbnail, captions, and queues it for the next available window: 7 AM, 3 PM, or 7 PM. Videos longer than four minutes get chopped automatically, with the original and the clips both distributed. Afro's diagnosis of why this matters is blunt: coaches can create content, but they cannot distribute it fast enough. If you record in the morning and it goes out in the afternoon, the bottleneck disappears. They are also running 31 active workflows, enough that Afro built a morning report just to track what is working and what has failed overnight. He is on the $100 a month Claude Max plan and was at 92% of his usage limit with five hours left in the billing cycle. He made it. For coaches who have never opened Claude, Afro's advice is direct: start with the free plan, be curious, and play. He recommends Claude over ChatGPT right now, specifically because it connects to Canva and other assets more cleanly, which matters for coaches building before and after photo content or curriculum slides. One student was building a course one slide at a time, a process Afro estimated could take six months. Claude stores it as a project, lets her update it by talking to it, and cuts that timeline sharply. His one firm boundary: if you have never coached a client and are starting from scratch, skip the agent work entirely. Agents replace and accelerate things that already exist. They do not build something from nothing. Get clients first, then automate the machine that serves them. Farah's line on household robots ("Does it cook dinners? Not yet. Wash dishes? Not yet.") is the only thing the technology cannot yet handle. Everything else, it seems, is already in the queue. This episode is for coaches who are already operating but feel buried in distribution, admin, and content logistics. The tools are here, the cost is $20 to $100 a month, and the barrier, as Afro puts it, is lower than it has ever been. Quick test for online coaches. Can you explain your coaching offer in one clear sentence? If that feels harder than it should be, you're not alone. Most coaches don't struggle because they're bad at coaching, they struggle because their offer is vague, hard to explain, or under-positioned. That's exactly why we created The Offer Room. It's a free community where online coaches come to clarify their coaching offer, sharpen their messaging, and structure premium offers they can confidently charge $2k+ for. You can join us here: https://www.skool.com/the-offer-room
