ALEX WORKSHOP SERIES

Hands-On Exercises

Practice by Discipline — Find Your Track · 10 minutes · 3 exercises

Hands-On Exercises

Welcome!

These exercises walk you through your first conversation with Alex. You’ll need:

Time: 10 minutes total | Exercises: 3

How to Use This Sheet

Exercise 1 is identical for everyone.

Exercise 2 has parallel tracks — scan the labels below and find your discipline. Do your row. The structure is identical across all tracks; only the examples change.

Exercise 3 is identical for everyone, with discipline-specific examples to choose from.

Discipline tracks available: Academic / Research · Engineer (Electrical, Mechanical, Civil, Process) · Developer (Software) Business Analyst / Knowledge Worker · Project Manager · Marketing · Content Creator · Student

Bring a real task from your current work. You’ll get far more out of the exercises with real context than a hypothetical.

Exercise 1: Meet Alex (3 minutes)

Setup

  1. Open VS Code
  2. Press Ctrl+Shift+P and type “Alex: Initialize Architecture” → press Enter
  3. Wait for the notification confirming initialization (~10 seconds)
  4. Look at the sidebar — you should see Alex’s Welcome View

Your First Message

  1. Open the Copilot Chat panel (click the Copilot icon in the sidebar, or press Ctrl+Shift+I)
  2. Type this — replacing the brackets with your information:
@alex Hello! My name is [your name]. I'm a [your role] working in [your field].
I prefer [brief/detailed] explanations and a [formal/casual] tone.

Example:

@alex Hello! My name is Maria. I'm a research analyst working in healthcare policy.
I prefer detailed explanations and a formal tone.
  1. Read Alex’s response. Notice:
    • Alex greets you by name
    • Alex may check its own health status
    • Alex remembers your preferences for future conversations

Verify

Type: @alex What do you know about me?

Alex should recall what you just told it. This is persistent memory — it will remember next time too.

Exercise 2: Dialog Engineering (4 minutes)

Use the Context-Goal-Constraints Pattern

Structure every substantive request as:

@alex I'm a [role], working on [real task or project].
I need [specific deliverable].
Keep it [constraint: length, format, audience, tone].

Find your discipline track below. Use the example as a starting point — or substitute your own real task.

Academic Researcher / Professor

@alex I'm a researcher preparing a literature review on [your topic].
I need help identifying the key theoretical frameworks and how they relate to each other.
Structure it as an annotated outline, not prose. Focus on the last 10 years of work.

Iterate:

Good. Now add a gap analysis — what does the current literature fail to address?

What would a reviewer's most likely criticism of this framing be?

Engineer (Electrical, Mechanical, Civil, or Process)

@alex I'm a [mechanical/electrical/civil/process] engineer working on [your project or task].
I need to document [design decision, selection rationale, or review preparation].
Format it as a structured technical justification referencing the relevant standard
([ASME/IEC/IEEE/ISO/local code as applicable]).

Iterate:

Add a section covering the failure modes we need to address with this design choice.

What would a code reviewer or regulator most likely flag in this document?

Developer (Software)

@alex I'm a software developer working on [your current system or feature].
I need an Architecture Decision Record (ADR) covering [the decision at hand].
Include context, options considered, decision, and consequences.
Use the standard ADR format.

Iterate:

Add a section on what we'd need to do to reverse this decision if it turns out to be wrong.

What are the non-obvious long-term maintenance costs of this approach?

Business Analyst / Knowledge Worker

@alex I'm a business analyst preparing a [report type] for [audience].
I need a framework for presenting [topic or situation].
Format it as [table/bullets/one-pager] — audience is [non-technical/executive/operational].

Iterate:

That's good, but cut it in half. Keep only the three most important points.

What's the most likely pushback I'll get from this audience? How do I pre-empt it?

Project Manager

@alex I'm a project manager running [initiative name or type] with [team size or composition].
I need a [communication plan / risk register / stakeholder map] for this project.
Format it as a table. Include [specific columns relevant to the deliverable].

Iterate:

Highlight the three highest-risk items and suggest mitigation actions for each.

If I had to cut the scope by 30%, what would you recommend I deprioritize and why?

Marketing

@alex I'm a marketing professional working on [campaign, launch, or content initiative].
I need [a campaign brief / content strategy / messaging framework] for [audience/product].
The goal is [awareness/conversion/retention]. Tone: [brand voice descriptor].

Iterate:

Good. Now write the hook for the first social post — 280 characters, same tone.

What would the biggest critic of this messaging argue? How do I address it?

Content Creator

@alex I'm a content creator working in [your niche or platform].
I need [a structure / outline / angle] for [video, article, newsletter, podcast episode].
Audience: [describe your audience briefly]. Goal: [inform/entertain/convert].

Iterate:

Give me three alternative angles on the same topic — more contrarian, more personal, more data-driven.

What's the most boring version of this idea? Why does my current angle avoid that?

Student

@alex I'm a student working on [assignment, research paper, or project] for [course].
I need help structuring my argument before I start writing.
The assignment is: [paste or summarize the prompt].

Iterate:

Good outline. What's the weakest part of this argument and how do I strengthen it?

What counterargument would a grader most likely raise? How should I address it?

The Observation

Three turns. No restarts. Each turn built on the last. That is Dialog Engineering.

A single prompt cannot do what three iterative turns can do.

Exercise 3: Save Your First Insight (3 minutes)

Save Something You Learned

Think of one thing you learned today — about Alex, dialog engineering, or AI in general. Save it:

@alex /saveinsight title="[short title]" insight="[what you learned, in your own words]" tags="workshop,alex,dialog-engineering"

Find your discipline below for an example of a strong, specific insight:

Academic / Researcher:

@alex /saveinsight title="Structure before synthesis" insight="Asking Alex to outline the key frameworks in a literature area before writing any prose produces a better review structure than asking for prose directly. The iterative approach surfaces gaps I wouldn't have noticed." tags="workshop,research,dialog-engineering"

Engineer (Electrical/Mechanical/Civil/Process):

@alex /saveinsight title="Design justification scaffolding" insight="Alex can structure a code-compliant design justification document faster than starting from a blank template — as long as I specify the standard and failure modes upfront. Always specify the standard in the first message." tags="workshop,engineering,standards"

Developer (Software):

@alex /saveinsight title="ADR iterative refinement" insight="Starting an ADR with Alex and then asking what the reversal plan would look like consistently surfaces risks I hadn't considered. The second and third turns matter more than the first." tags="workshop,architecture,adr"

Business Analyst / Knowledge Worker:

@alex /saveinsight title="Iterate instead of restart" insight="When working with AI, refining a response in 2-3 turns produces much better results than restarting with a new prompt each time. The AI uses the accumulated context to improve." tags="workshop,alex,dialog-engineering,productivity"

Project Manager:

@alex /saveinsight title="Risk register prompt pattern" insight="Giving Alex the stakeholder list, initiative goal, and team size upfront produces a risk register that needs only 20% editing. Without that context, the output is generic." tags="workshop,pm,risk,dialog-engineering"

Marketing:

@alex /saveinsight title="Campaign brief context window" insight="The more specific the audience description and the explicit tone constraint in the first message, the less time I spend editing AI marketing copy. Brand voice always goes in the first message." tags="workshop,marketing,brand-voice"

Content Creator:

@alex /saveinsight title="Multiple angle generation" insight="Asking Alex for three angles — contrarian, personal, data-driven — on the same content idea breaks creative blocks faster than asking it to write something interesting about X." tags="workshop,content,creativity"

Student:

@alex /saveinsight title="Pre-writing structure check" insight="Asking Alex to review my outline before writing the paper and identify the weakest argument saved more time than asking it to review the completed draft. Fix structure early, not late." tags="workshop,writing,student"

Search For It

Now search your knowledge base:

@alex /knowledge dialog engineering

Your insight should appear in the results. This insight is now stored in your Global Knowledge base and will be available in every project where you use Alex.

Check Your Knowledge Status

@alex /knowledgestatus

This shows your complete knowledge library — patterns, insights, categories, and contributing projects.

Bonus: If You Have Extra Time

Try Explain-Like

@alex Explain blockchain like I'm a business executive who understands databases
but has no background in cryptography. Use a real-world analogy.

Try Creating a Diagram

@alex Create a flowchart showing the decision process for approving a new project
in our organization. Include: proposal submission, budget review, technical review,
executive approval, and the possible outcomes at each gate.

Try the Rubber Duck

@alex I'm going to think out loud. Just listen and help me organize my thoughts.

I'm trying to figure out whether we should centralize our data analytics team
or keep them embedded in each business unit. The centralized model gives us
consistency but the embedded model gives us speed. We tried centralized before
and it created bottlenecks. But the embedded model has led to duplicate work
and inconsistent methodologies...

What pattern do you see in what I just described?

What’s Next?

After this session:

  1. Keep the dialog going — Use your real task from today for the next conversation with Alex
  2. Save insights regularly — /saveinsight is a 30-second habit worth building
  3. Meditate weekly — @alex /meditate consolidates everything you’ve learned
  4. Check your study guide — Your facilitator will send a discipline-specific guide with week-by-week application steps

“One workshop won’t change how you work. Consistent practice will.”