Skip to main content

Live Architecture Session

The signature segment of Applied AI Live. A real business owner presents a real problem. A practitioner architects a solution on the spot. The audience learns by watching.

This guide covers how to find the right business owner and prepare the engineer for success.


Finding the Right Business Owner

You do not want to pick someone random from the audience. You want someone pre-vetted with:

  • A real business that is earning revenue
  • A genuine desire to level up their operations
  • Openness to AI integration (not skeptical or resistant)
  • Willingness to have their problem workshopped publicly

The best candidates are people you already know. Ideally, they're friends or close contacts who trust you enough to be a guinea pig. They understand this is experimental and valuable for everyone involved.

What Makes a Great Candidate

Look for people who love their work broadly but feel stuck doing too much of the wrong kind of work.

There's a framing that resonates with most business owners:

There's a bucket of work you're dying to get to. The stuff that puts you in flow state. The work that actually moves the needle. And then there's the other stuff. The infrastructure. The admin. The things that make the business run but drain your energy.

Every business has this tension. The best candidates feel it acutely. They're not looking to "automate everything." They want to protect the parts that require their soul while offloading the parts that don't.


The Soul Work vs. Soulless Work Frame

This is the pitch. Adjust the language for your audience, but the structure works:

Soul Work: The stuff that requires you. Your taste, your judgment, your relationships, your creative vision. This should never be automated.

Soulless Work: The infrastructure that keeps things running. Accounting, scheduling, contract generation, logistics, lead qualification. This can and should be delegated or automated.

When you talk to potential candidates, help them articulate what falls into each bucket. If they light up when describing the soul work and sigh when listing the soulless work, they're a good fit.

Example Categories

Soul Work (Keep Human)Soulless Work (Automate)
Creative vision and tasteAdmin and scheduling
Client relationshipsAccounting and expense classification
Strategic decisionsContract drafting and review
Collaboration and communicationLogistics and coordination
Final product judgmentLead research and qualification

Outreach Template

Here's a message you can adapt when reaching out to potential candidates:

Hey [Name],

I'm organizing an event called Applied AI Live where practitioners help business owners solve real problems using AI.

There's a segment where one business owner gets to workshop a problem live with an experienced applied AI practitioner. The practitioner architects a solution on the spot while the audience watches and learns.

I immediately thought of you. You have a real business, you're thoughtful about how you work, and I think there's a lot we could dig into.

The way I'd frame it: there's probably work you're dying to spend more time on. The stuff that puts you in flow. And there's work that gets in the way of that. We'd explore where AI could handle the second category so you get more time for the first.

Would you be open to being our first guinea pig? We'd do a quick call beforehand to map things out, then I'd write up a brief for the engineer so the session is tight and useful.

Let me know if you're interested.


The Pre-Call

Before the event, have a 30-60 minute conversation with the business owner. The goal is to understand their business deeply enough to write a problem brief.

Recording the Call

With permission, record and transcribe the entire conversation. Then feed the transcript plus this playbook to an AI (Claude works well) and have it generate the problem brief. This saves you from manual note-taking and produces a more thorough brief.

Any AI notetaker works. The key is capturing everything so you can focus on the conversation, not documentation.

What to Cover

  1. Business overview: What do they do? Who are their clients? What's their revenue model?

  2. Soul work vs. soulless work: What energizes them? What drains them? Where do they spend most of their time vs. where they want to spend it?

  3. Current pain points: What's not working? What takes too long? What falls through the cracks?

  4. What they've tried: Have they used AI tools already? What worked? What didn't?

  5. Constraints: Solo operator or team? Budget? Non-negotiables (e.g., "relationships must stay human")?

  6. Success criteria: What does "better" look like? How would they know if AI was helping?

Take notes. Ask follow-up questions. Get specific examples. The more concrete the brief, the better the live session.


Writing the Problem Brief

The problem brief is a document you give to the practitioner at least one week before the event. It's like a hackathon challenge prompt. They need time to think through the problem before going on stage.

Brief Structure

  1. Business Overview: Who is the client? What do they do? Who are their customers?

  2. Core Offerings: What products/services do they provide?

  3. The Business Problem: What's broken? What's the tension between soul work and soulless work?

  4. Current Pain Points: Specific issues, listed clearly.

  5. Constraints: Budget, team size, non-negotiables.

  6. What They've Already Tried: AI tools, systems, workarounds.

  7. Architecture Challenge: The core question for the engineer to solve.

  8. Specific Use Cases: Concrete workflows that could be automated or augmented.

  9. Success Criteria: How the business owner defines success.

  10. Questions for the Live Session: Prompts to guide the on-stage discussion.

Tone

Write it like a case study. Clear, specific, actionable. Avoid jargon. Include direct quotes from the business owner when possible.


Preparing the Engineer

The engineer going on stage needs:

  1. The problem brief: At least one week in advance.

  2. Context on the format: They'll have ~30 minutes to architect a solution live. Whiteboard or large sticky notes. Audience will be watching.

  3. Permission to ask questions: The session isn't a lecture. The engineer should dialogue with the business owner on stage.

  4. A heads-up on constraints: What's in scope? What's off limits? Any sensitive topics?

Give them space to prepare however they want. Some engineers will sketch solutions ahead of time. Others will prefer to think on their feet. Both work.

What You're Looking For

A good live architecture session:

  • Clarifies the problem in plain language
  • Proposes a realistic solution (not science fiction)
  • Considers tradeoffs and constraints
  • Gives the business owner something actionable to walk away with
  • Teaches the audience something they can apply to their own work

Example Brief

Applied AI Live #1 (January 2026) used this format successfully. The session held the room and received strong positive feedback. A full example brief will be published separately. In the meantime, use the structure above as your template.


Checklist

Before the event:

  • Business owner confirmed and briefed
  • Pre-call completed (30-60 min)
  • Problem brief written
  • Brief sent to engineer (1 week before)
  • Engineer confirmed and prepared
  • Whiteboard or sticky notes available for the session
  • Business owner knows what to expect on stage

Why This Matters

The live architecture session is the most distinctive part of Applied AI Live. It's not a demo. It's not a pitch. It's real problem-solving in front of an audience.

When done well, it:

  • Gives the business owner a genuine head start on their AI journey
  • Shows the audience what applied AI work actually looks like
  • Builds credibility for the engineer
  • Creates content worth talking about after the event

The prep work makes the magic possible. Don't skip it.

At Applied AI Live #1, the live architecture session was the standout segment. The business owner's problem (and the "Sovereign AI Command Center" concept the engineer proposed) resonated with multiple attendees in post-event conversations. One lesson: brief the architect more thoroughly in advance. The more context they have, the tighter the session.

Live architecture diagram from Applied AI Live #1