Solutions StaffRoomAI Training

Unit 1: Conducting the Conversation

Stop thinking in “prompts I send.” Start thinking in “a working session I’m conducting.” Unit 1 installs the foundational mental model every subsequent unit builds on—anchored to real pieces of your work.

Built for experts, not for novices

You already know how to run a working session. You brief associates, redirect clients, push back on stagehands, debrief with surgical teams, coach students. That instinct is exactly the skill needed here. The barrier isn’t capability—it’s recognizing that this is that, pointed at a new kind of collaborator.

Unit 1 makes that movement conscious so it can become unconscious again—which is what fluency actually is.

What you’ll learn

Three concepts, each anchored to a real piece of your work.

The conversation is the unit of work

From “I write a prompt and judge the output” to “I conduct a working session and shape it as it unfolds.”

Experts already know how to run working sessions—with associates, clients, students, surgical teams. The skill is more than enough; it just needs to be pointed at a new kind of collaborator. You learn the working loop, when to edit vs. retry vs. respond, and how to treat a 70% draft as material to refine.

Context is what you bring

From “What features does Claude have?” to “What does this collaborator need to see to be useful to me?”

Before any task, ask: if I were briefing a smart colleague who has never seen this work, what would I put in front of them? You learn to stage context deliberately—files, images, search, connectors—and, just as important, to recognize when less context produces better output.

Persistence is a design choice

From “Should I make a project?” to “Does this work deserve a persistent collaborator, and if so, how should that collaborator be shaped?”

Recurring contexts deserve persistent setups—but only when recurrence, shared context, or consistent voice justify it. You learn to use instructions as onboarding, treat memory as reviewable rather than magical, and maintain projects the way you maintain any professional system.

How the three layers connect

These aren’t three modules. They’re three layers of the same thing, each operating at a different timescale.

Minutes

Conversation

This working session

Per task

Context

What this session can see

Across sessions

Persistence

Who the collaborator is for you

A fluent expert moves between all three layers without thinking about it.

What you’ll walk away with

  • Articulate a real working problem to AI in plain language—without translating it into “prompt format.”
  • Recognize when to edit, retry, or respond mid-dialogue, and use each appropriately.
  • Treat first-draft output as material to evaluate, not a verdict to accept or reject.
  • Decide before prompting what context the AI needs, and stage it deliberately.
  • Match the right context channel to the situation—file, image, search, or connector.
  • Identify at least one recurring context worth persisting as a project, and one task that should stay a one-off.

What Unit 1 does not do

The framework is anti–feature-tour. So it’s worth being explicit about what we leave out.

  • Model selection, settings, or advanced features. Those are tool-layer details and would recreate the software-tutorial trap.
  • The seven problem categories (synthesis, decision support, etc.). That is Unit 2’s job—Unit 1 just establishes that a relationship exists and you shape it.
  • Prompt patterns or templates. Your personal pattern library is built through practice across many units, not handed out as a starter pack.

Join Unit 1

Sign up below and we’ll be in touch with the next cohort dates and what to bring.

Want to bring StaffRoomAI Training to your team?

We run Unit 1 (and the rest of the program) for whole organizations too. Book a strategy session and we’ll scope it to your context.

Book a Free Consultation
StaffRoomAI Training — Unit 1: Conducting the Conversation | StaffRoomAI