How lecturers use knowledge management processes to turn lesson preparation into a reusable system

What is this article talking about?

This article lays out the knowledge management workflow behind the "Lecturer's Agent Workflow." It starts from making presentations and extends to how a lecturer can turn everyday reading, student questions, community feedback, pre-class questionnaires and post-class transcripts into material that can be pulled up directly when preparing the next class.

The eight-stage workflow: from material to a reusable system

This class breaks the lecturer's workflow into eight stages. Each stage produces knowledge assets you can reuse next time, not just something that serves the task at hand.

Finding information: don't let one AI search in a single direction

AI can easily get pulled off course by context when it looks for information. If you ask it about the reception of some product launch, it may find the official material first, and then let the official framing steer the whole line of reasoning.

The approach taught in the course is to run multiple agents in parallel: have different sub-agents separately find the official material, the positive takes and the negative takes, then pull the sources and original excerpts back for cross-comparison. The AI doesn't draw the conclusion directly; its job is first to gather the material and list the sources clearly.

This matters a lot for a lecturer, because the biggest risk in lesson prep is that you only see the material you already believe, and then package that bias into your course. The value of running agents in parallel is that you get to see different angles before you prepare.

Sorting, categorizing and the inspiration pool

A lecturer often sees a good example, hears a student question, or reads about a trend without an immediate use for it. But if you don't note it down in the moment, it's hard to recall when it's actually time to prepare.

The class frames this as three actions: accumulate as you go, thread with tags, and summarize on a schedule.

Accumulating as you go isn't just saving the content; it's also noting why it struck you as useful at the time. Threading with tags uses general tags and session tags to connect material to courses you might teach later. Scheduled summarizing sets a rhythm so material doesn't pile up too long before it gets organized.

Pre-class questionnaire: turn "what do you want to learn" into "which step are you stuck on"

Many pre-class questionnaires ask, "What do you want to learn?" But students sometimes don't know what they want to learn.

What the course values more is asking, "Which step are you stuck on right now?"

This way of asking is closer to genuine needs research. Students may not be able to name the full solution, but they usually know where they're stuck. Once the lecturer has the responses, they can use an Agent to run a placement analysis and see which levels of ability and which kinds of needs cluster together.

This changes how you prepare. Before designing the course, a lecturer can first see where this group is really stuck, then go back and calibrate their own assumptions. The core section handles what most people struggle with first; a supplementary section handles the few extended needs.

Pre-class promotion is part of lesson prep too

Promotion before a course isn't only about enrollment; it's also a way to test how your audience understands the topic.

The course walks through the workflow for social carousel cards: write the copy first, then generate the storyboard, and only then create the images. The copy is the backbone, the storyboard is the bridge, and the images are the flesh.

The knowledge management point behind this workflow is that every round of promotion generates feedback. When someone comments asking, "Will you cover such-and-such?" that kind of signal can become lesson-prep material. It can go back into the inspiration pool, or it can change how you open the actual course.

If a lecturer has a fixed brand character or visual style, like Mika, then the style definition, three-view reference and go-to storyboarding principles should be saved too. This material keeps the next round of visual content consistent.

Lesson preparation: use tags to gather material into one bundle

When it's actually time to prepare, the key is to gather the material you've already tagged along the way, so you don't have to start from a blank page.

The approach shown in the course is this: on prep day, open your lecture index page and use backlinks or tags to pull out the relevant material, such as "Lecturer Workflow" and "2026-05-23 Course." That material might be work diaries, half-finished articles, student feedback, community comments, or the Q&A from the previous class.

Then let AI polish the draft. AI can handle tone, rhythm and transitions; the lecturer stays responsible for the stance and the judgment calls.

Making deliverables: keep the finished product saveable, shareable and AI-readable

Deliverables can be slides, web presentations, after-class handouts, example files, prompts, or a complete teaching manual.

The course uses a web presentation as its demo, with the focus on deliverables being saveable, shareable and readable by AI. That's exactly why HTML presentations are worth it: they can be projected, and they can also become material that the next Agent can search, recycle and reuse.

Public content can go on GitHub Pages. Anything containing meeting records, client reports, personal data or sensitive content should go through semi-private deployment or stay unpublished.

Post-class repurposing: keep the details, not just a summary

What really matters after class is to save the details of what happened in the room, instead of only writing a summary.

The class covers a three-level approach to organizing transcripts: super-detailed, very detailed, and 1:1. The value isn't only in organizing content but also in the insight layer. You can have AI analyze where you spoke smoothly and where you didn't, which passages students responded to well, and which parts need reworking.

After class you can also build a teaching manual, turning the lecturer's own notes into a version someone else could pick up and teach. Student questions can be filed one by one and linked back to the index page for the next class on the same topic.

That way, student questions carry over from classroom interaction into the best material for your next round of prep.

Cross-course reuse: turn courses into building blocks

Once all seven earlier stages are kept, the lecturer no longer has to start from scratch every time.

One successful run can be written into a skill package; a stable way of explaining something can be broken into cards; a batch of student questions can become the FAQ for the next course; a transcript can turn into a teaching manual or an in-depth article.

The course uses building blocks as its metaphor: card-based knowledge can be rearranged and recombined to fit the needs of the next class. The same body of knowledge can be presented differently to university professors, administrative staff, content creators, social media managers or business executives, while the underlying material stays shared.