Two-Phase Transcript Analysis

In this episode, Aydin sits down with Ryan McCready, who went from hating AI to becoming one of the most creative AI-powered content builders on the internet. After getting laid off in mid-2025, Ryan realized that every job interview demanded AI fluency. So he went all-in, teaching himself Zapier, Lovable, Supabase, and advanced prompting to engineer a “Content Factory” that turns a webinar into blog posts, clips, and social content in minutes.He shares the mindset shift from “AI is plagiarism” to “AI is an input-output engine,” why content engineering is the future, what makes AI workflows actually work, and how breaking big tasks into many small steps is the secret to non-sloppy AI content.You’ll see how he built a 30-step Zapier workflow that analyzes a webinar transcript, extracts frameworks and insights, turns them into pitches, builds outlines, writes social posts, and even generates clip candidates for Descript. If you create content or run marketing—this one is a masterclass.Timestamps0:23.00 – Why he believed AI was a “plagiarism machine”2:04.00 – Getting laid off and realizing every employer wanted AI skills4:37.00 – The workflow that kickstarted his learning (LinkedIn voice extraction + employee advocacy shares)5:40.00 – Learning Lovable and Supabase by building real projects6:51.00 – Why “everyone is a builder now” because of AI tools7:52.00 – Introducing “Content Engineering” and why most marketers can’t do it9:03.00 – Example: turning a webinar into 10+ pieces of content10:58.00 – Why webinars usually die after they’re aired—and why that’s a waste11:43.00 – The “Webinar Content Flywheel” teaser16:30.00 – Why Ryan moved back from n8n to Zapier17:55.00 – Zapier vs. n8n: simplicity, stability, and architecture19:03.00 – “Start small”: a two-step Zap example20:09.00 – Interface demo: uploading a transcript and hitting “Go”21:22.00 – Why Zapier Interfaces make deployment easy22:40.00 – Step-by-step breakdown of the workflow24:06.00 – Example: webinar analysis output (themes, chapters, frameworks)27:02.00 – Creating three blog pitches from the transcript30:43.00 – Sending the pitches to Slack for review31:03.00 – Clip extraction workflow + Descript integration32:14.00 – How he uses Descript’s “Underlord” to auto-cut clips33:20.00 – Why this beats automated clip tools like Riverside for B2B35:02.00 – Social content workflow (framework angle, data angle, hot take, wildcard)37:12.00 – Why prompting manually is wasteful—build once, automate forever40:11.00 – “Big → small → big” framework: the secret to non-sloppy AI content41:21.00 – Google’s “AI content penalty” myth, according to Ryan42:47.00 – Why your input quality determines whether your AI output is good43:44.00 – What excites him most in the next 12 monthsTools & Technologies MentionedZapier: Automation platform used to chain 30+ steps together: analysis, pitch creation, clip extraction, social content, Notion updates, etc.AI by Zapier: Zapier’s built-in LLM module used for analysis, extraction, outline generation, and writing.n8n: Open-source workflow automation platform. Ryan tested it, but ultimately moved back to Zapier for stability and structure.Lovable: AI-enabled “vibe coding” tool that turns prompts into functional web apps.Supabase: A database + backend platform used for storing structured content data from builds.Descript (Underlord): Video editing tool with an AI agent that cuts clips based on transcript timecodes generated by the workflow.Notion: Used as the source of truth for storing transcripts, outlines, clip docs, and the full content tracker.Claude / ChatGPT: Used for second-pass expansion—turning outlines or social angles into fully polished blog posts and posts.Fellow.ai: AI meeting assistant—summarizes meetings, tracks decisions, and generates insights and performance summaries.Subscribe at⁠ thisnewway.com⁠ to get the step-by-step playbooks, tools, and workflows.