Anyone else having issues with Retell and the cal.com integration today? It's been working fine, but this afternoon every attempt by the agent to actually schedule a time fails. The check_availability seems to work properly, but the voice agent continually says it's run into an issue with the system when trying to actually book the call.

Oh man, I needed this win. 😅 WIN: After-hours AI routing system finally LIVE for a real restoration company 🎉 I’ve been in the trenches for MONTHS trying to get a real production call-routing system working for a national restoration company … and tonight it finally passed a full live test end-to-end. 🙌 What I just shipped: An after-hours emergency routing system where: - A Retell AI agent answers the emergency line - Collects caller info + ZIP code - Calls a custom tool that hits a Relay.app webhook - Relay talks to multiple Google Sheets to figure out: Is this weeknight vs weekend? Who’s on call (PC vs PM)? Which region (East vs West)? - - Relay sends back the exact human + phone number - Retell does a warm transfer straight to the on-call person I tested it live and the call actually rang the correct on-call contact. No manual fix, no “almost there”. It worked. 😭🔥 Why this feels huge: - I almost abandoned this because I hit so many walls. - Claude’s rate limits + weird errors slowed me down hard. I ended up abounding Claude and using Chat GPT and it did fantastic. AND NO STUPID RATE LiMITS! - I kept breaking little pieces in Relay — path logic, JS step outputs, missing fields, you name it. - There were plenty of “I’m going to shut this all down” moments. But I stuck with it, untangled the logic one step at a time, and now: - Retell + Relay + Google Sheets are working as a real, revenue-generating system - The business now has a 24/7 emergency line that routes intelligently instead of burning out staff - I have a repeatable pattern I can clone for other service businesses (restoration, HVAC, plumbing, etc.) I just wanted to share this because it’s one of those “this is actually running in the real world now” milestones that’s easy to skip past and move on from. If you’re stuck in the middle of a gnarly build: keep going. Your future “live test passed” moment is ridiculously satisfying. 💪 What’s a recent win you’ve had, even if it felt small at the time?

Just wrapped up another automation project, and to be honest, this one felt really good to deliver. Steven came to me with a familiar problem, he was spending too much time doing the same repetitive tasks every single day. Nothing too crazy, but enough to slow him down and drain his focus. His words were basically: “I just want this stuff to run without me babysitting it.” So I built him a workflow that kind of does exactly that. Simple input → clean automation → consistent results No stress. And seeing feedback like this afterwards… honestly, it means a lot: ⭐️⭐️⭐️⭐️⭐️“Johnson has been really professional and very knowledgeable about the subject matter and it’s reflected on the automation workflow he created for me. I would recommend him to anyone.” I think this is why I enjoy building these systems, people get their time back, their work gets smoother, and the annoying parts disappear in the background. If you could automate one thing in your daily workflow, what would it be?

Client paid their agency $31k over 4 months for "workflow optimization." I recreated everything from scratch in 67 minutes. Runs 4x faster. Sometimes starting over beats fixing broken things. THE PROBLEM Client: "Our workflows are getting slower" Agency: "You need our optimization package" Client: "What's that involve?" Agency: "Performance audit, code refactoring, architecture improvements" Cost: $31,000 Timeline: 4 months Month 1: Audit phase Month 2: Planning phase Month 3: Implementation Month 4: Testing and rollout Client paid. THE RESULTS After 4 months and $31k: - Workflows 12% faster - Three new bugs introduced - Monthly maintenance required - Additional $2k/month ongoing support Client called me: "This doesn't feel right" THE ANALYSIS Me: "Show me the workflows" Client shares access. I review their system. What I see: - 9 workflows - Overly complex routing - Redundant steps everywhere - Built like spaghetti code Me: "Your agency optimized this?" Client: "That's what they said" Me: "They polished a mess" THE SOLUTION Me: "Let me rebuild these from scratch" Client: "How long?" Me: "Give me an hour" Client: "For the estimate?" Me: "For the rebuild" THE REBUILD 1:00 PM - Open Skada.ai Workflow 1: Lead processing - Agency version: 24 nodes, 8 seconds - My version: Generated in 4 minutes, runs in 1.8 seconds Workflow 2: Order fulfillment - Agency version: 31 nodes, 12 seconds - My version: Generated in 6 minutes, runs in 2.4 seconds Workflow 3-9: Same pattern 2:07 PM - Done All 9 workflows rebuilt. All tested. All deployed. THE PERFORMANCE Agency's "optimized" workflows: - Average execution: 9.2 seconds - Error rate: 3.1% - Maintenance: Weekly updates needed My rebuilt workflows: - Average execution: 2.1 seconds - Error rate: 0.1% - Maintenance: None needed 4.4x faster. Built in 67 minutes. THE COMPARISON Agency approach: - Analyzed existing complexity - Added patches and fixes - "Optimized" inefficient structure - Charged $31,000 - Took 4 months

We’re looking for an AI-first developer who is great at using LLMs and AI tools to build real products. You do not need to be a “10x traditional coder” — you need to be excellent at orchestrating AI (Claude, ChatGPT, Gemini, Copilot, Cursor, etc.), turning ideas into working features, and iterating quickly. You’ll help us prototype and ship features inside AX Platform, build integrations and workflows on top of MCP, and use AI as a core part of your daily development process.https://ax-platform.com/careers/ai-developer/