Level 2 is where your whole system stops being “creative effort” and becomes repeatable throughput. If Level 1 is the Content Engine (idea → lesson kit), Level 2 is the Compression Layer: the workflows that take that lesson kit and turn it into finished assets quickly, consistently, and in your voice.
The key here is that visuals are not decorative; they’re script-anchored. Each diagram corresponds to a specific explanation beat, so your lesson stays coherent across video, slides, and LMS.
Pipeline B: Dialogue + avatar production (Meg & Baz workflow)
When the lesson benefits from conversational friction, you generate scripts in a defined style using Claude or ChatGPT, producing dialogue between you and the AI character (often Meg).You generate Meg’s base image in Midjourney, then refine her clothing/background in Gemini to match the episode’s setting.
You then upload the image and script into Heygen to output the Meg video segment, including translation when needed. In parallel, you film your own segment for your part of the dialogue, then assemble both halves together in DaVinci Resolve into a single cohesive lesson video.
The workflow starts with a simple principle: don’t use AI as a magic trick — use it as a production line. You begin with a stable input (your session script or lesson outline), then you run that through a sequence of “transforms” that each produce a specific output you can file, reuse, and publish. In your outline, the first transform is script polishing: you research, draft, then use AI to polish the session script so it becomes clean, spoken, and structured for delivery.
From that polished script you branch into two parallel asset pipelines:
Pipeline A: Visuals (images + keypoint charts)
You extract image prompts from the script and feed them into Midjourney to generate marketing and lesson imagery. In parallel, you extract “infographic / flowchart / keypoint chart” prompts and generate those in Gemini (“Nano banana”), where you’ve already configured style presets so the charts come out with consistent formatting and look-and-feel.
Once assets exist, Level 2 adds the scale mechanism: automation and orchestration. You use n8n to combine code, AI agents, and system messages into automated workflows, triggered through Slack commands, with inputs/outputs logged into Airtable for traceability and reuse.
You can also push content into Pinecone and query it through n8n, while keeping research in NotebookLM so it remains searchable over time like a living knowledge base.
By the end of Level 2, learners aren’t just “using tools.” They’ve built a content compression workflow: script in → prompts out → visuals + avatar segments → edit assembly → publish-ready asset bundle, with automation that reduces rework and makes output predictable.
Release Schedule
The Four Levels Pathway
This series is being released as a four-level pathway, because building a content and production capability isn’t something you “learn once” — it’s something you layer. Each level gives you a complete, usable capability on its own, and each release adds the next layer of power without forcing you to rebuild what you already have.
The schedule is designed around a simple idea: different people are starting from different places. Some of you already have a website, a course platform, and an audience — your constraint isn’t motivation, it’s throughput. You don’t need a studio tour; you need a faster, cleaner way to turn your thinking into publishable assets, using AI and workflow design.
Others want the opposite: you already have content ideas, but you want to deliver with confidence — a setup that lets you run workshops, record modules, and switch between camera, slides, demos, and examples like it’s normal. And some of you want the whole machine: content that’s coherent, production that’s reliable, and a publishing system that compounds week after week.
So the release schedule is staged in the same sequence the capability is built: