AI Starter Map — Webinar Runbook

01 · for attendees

Program

What attendees see — share-friendly. Copy verbatim for the landing page, registration confirmation email, and pre-event reminder.

What you'll get

Two live sessions, a week apart, designed to take you from “AI sounds important but I don't know where to start” to “I have a working prototype I can show my team.”

→A method to identify your most valuable AI opportunity in under 60 minutes.
→A worksheet you'll fill in live (Friction Map + Time × Impact matrix + Flow Designer).
→A live build: we'll take a real attendee's flow and prototype it in 25 minutes using no-code tools.
→Recordings of both sessions, kept evergreen.

Session 01Monday — Day 1· 60 min · live

“Find your AI starting point”

You'll leave with a prioritized quick win and a designed flow — on paper, before any tools.

5 min

Hook + intro

Why most AI initiatives never reach production — and what to do instead.

5 min

3-step framework

Map of the method we will cover across both sessions.

15 min

Friction Map

List the 10 most repetitive, costly, error-prone tasks in your operation.

10 min

Time × Impact matrix

Identify your one quick win — high impact, low effort. Avoid the "strategic" trap.

15 min

Design the flow

Pen and paper first: trigger → steps → output → metric.

10 min

Pre-close + Q&A

Homework: bring a finished flow to session 2 for a chance to see it built live.

You leave with

A worksheet you can re-use across your team: one prioritized quick win + one fully-designed flow.

Session 02Next Monday — Day 8· 60 min · live

“Prototype it in under 2 weeks”

One attendee's flow gets built live in 25 minutes using no-code tools. You see the method in action, errors included.

5 min

Recap + warm-up

Reactivate the worksheet from session 1.

10 min

Prototype mindset

Why no-code first. When to pick Zapier vs Make vs n8n vs custom.

10 min

Failure case study

The "Sofía" story: a prototype that broke production but validated the hypothesis.

25 min

★LIVE BUILD

Tom builds a real flow from the audience using Zapier + OpenAI + WhatsApp.

5 min

When to scale

Signals that a prototype is ready for proper build.

5 min

CTA + close

For wellness operators: applications open to AI Blueprint Webinar A.

You leave with

A working reference prototype you can adapt to your own flow, plus a mental model for prototyping the rest in under two weeks.

before session 1

Pre-session checklist

▸Download the Friction Map worksheet (sent 3 days before).
▸Block 60 minutes — uninterrupted. The activities only work if you actually do them.
▸Camera optional. Chat is where the action happens.

apply now

Two Mondays. One method. A working prototype.

Seats are limited. We pick attendees who'll actually do the work.

Apply to the series See full agenda

02 · for tom

Facilitator minutes

Run-of-show with timing, speaker prompts, activity instructions, and contingency notes. Private.

Private — Tom onlydo not share with attendees

Session 01 run sheet60 min · live

“Find your AI starting point”

backbone — "You don't have an AI problem. You have a friction problem."

0–5'

Hook + intro

say

›70% of SME AI initiatives never reach production — not because of tech, because of where they start.
›Anchor: AI doesn't replace humans. It replaces inefficient processes.
›Self-intro: CTO Amalgama, ~14 years, wellness ops focus, 30+ products in production.

›Slides 1–4 (title / stat / anchor / Tom intro).

5–10'

3-step framework

say

›Today: steps 1 + 2. Next Monday: step 3 — live build.
›Common failure: jump straight to "what tool", skip the diagnosis.

›Slides 5–6 (framework + what we cover today vs next week).

POLL

Where are you stuck right now?

I don't know where to start
I tried and it failed
I have pilots but they don't scale
I'm just exploring

Share results live. If (a) wins, anchor: "Perfect, this session is for you."

10–25'

Friction Map

say

›Define: list of repetitive / costly / error-prone tasks across your op.
›Lenses: time lost, money lost, customers lost, quality inconsistent.
›Show case: real Amalgama wellness client with 30% weekly hours on X task.
›Trap: only thinking digital. Best frictions usually live in handovers between people.

›Slides 7–10. Show real before/after numbers from the case.

WORKSHEET

List your top 5 frictions (3 min silent).

Page 1 of the worksheet.

CHAT

Drop 1 friction in the chat — one sentence.

Read 3–5 aloud, look for patterns ("manual data entry" × 3 = signal).

25–35'

Time × Impact matrix

say

›Four quadrants: quick wins / strategic / minor / discard. Start with quick wins.
›Why: build momentum, generate case, learn the tools.
›Trap: smart founders fall in love with strategic. Fine — but AFTER the first quick win.
›Anchor: what 20% of focus drives 80% of the impact?

›Slides 11–14 (matrix + trap + anchor).

WORKSHEET

Place 5 frictions on the matrix (4 min).

Page 2 of the worksheet. Mark your top quick win.

CHAT

Share your top quick win in the chat.

Mentally pick the most interesting one — candidate for S2 live build (don't announce yet).

35–50'

Design the flow

say

›Pick ONE flow. Not five.
›Three non-negotiables: one flow / one owner / one metric.
›Pencil and paper first — tool second.
›Show the before/after of a real Amalgama flow (6 steps → 2 steps).

›Slides 15–19. Show the actual hand-drawn before/after.

WORKSHEET

Write your flow: trigger → steps → output → metric (5 min).

Page 3 of the worksheet.

CHAT

Optional: drop your metric in the chat.

Comment on quality — binary, countable, comparable before/after.

50–60'

Pre-close + Q&A

say

›Today you leave with: Friction Map + quick win + flow designed.
›Next Monday we take one real flow from the audience and build it live.
›Homework: finalize your flow, reply to tomorrow's email with it. Most interesting one is built live.
›See you Monday. (NO mention of AI Blueprint W-A yet — that's S2.)

›Slides 20–23.

Q&A

Q&A (5 min).

Prioritize method clarity. If technical → "Save it for next Monday, we are building live."

gap weekbetween S1 and S2

What happens in the 7 days

T+1 (Tue):Email out — recording S1 + worksheet PDF + CTA to send the flow back to Tom's inbox.
T+3 (Thu):Reminder email — “send your flow by Friday for a chance to be built live”.
T+4 (Fri):Tom reviews submissions. Picks 3 candidates. Confirms 1 with the author.
T+5 (Sat):Tom does a dry-run build of the chosen flow. Catches any blockers. Confirms Plan B pre-built.
T+6 (Sun):S2 reminder email — “see you tomorrow + bring your laptop”.

Drop-off risk: 30–50%. Mitigation = real homework + worksheet value self-contained in S1.

Session 02 run sheet60 min · live

“Prototype it in under 2 weeks”

backbone — "You don't need to build it. You need to validate it. Tools come last."

0–5'

Recap + warm-up

say

›Reactivate session 1: worksheet → quick win → flow.
›Rules today: hands-on. If something breaks live — that's the point.
›If you applied to the series, you've done 60% of the work. Today we finish.

›Slides 1–4.

POLL

Did you bring a flow today?

Yes, fully designed
Half-done, still thinking
No flow yet, just observing

If (c) wins, reframe: "We show you the method; you finish your flow this week."

5–15'

Prototype mindset

say

›Definition: smallest thing that validates the hypothesis.
›Why under 2 weeks: longer than that = you're still designing.
›Expensive trap: hiring a dev / custom build before validating.
›Why no-code first: speed, cost, reversibility, learning.
›Stack tour: orchestration (Zapier/Make/n8n), forms (Typeform), AI (OpenAI/Anthropic), storage (Airtable/Sheets/Notion).
›Decision tree: simple trigger→action = Zapier. Branching logic = Make/n8n. Generated text = LLM in the right step. Scale > 10k users = sales call.

›Slides 5–10.

POLL

Which of these tools have you used?

Zapier
Make / Integromat
n8n
OpenAI API directly
None of the above

Multi-select. Calibrates the live build narration.

15–25'

Failure case "Sofía"

say

›Hypothesis: bot that calls inactive users to bring them back.
›Initial build: simple prototype in N days.
›What went wrong: bot called every 20 minutes. Users hated it.
›What went RIGHT: the first calls did recover users.
›Lesson: the prototype failed as a system, but validated the hypothesis. That's success.
›Amalgama parallel: a real prototype of ours that failed but validated. (Pick the case in advance.)
›Rule: if your prototype can't fail, it's not a prototype. It's production.

›Slides 11–14. Show real numbers from the Amalgama case.

CHAT

What's the worst that could happen if your prototype goes wrong tomorrow?

Read 2–3. Most overestimate the cost. Anchor: inaction is more expensive than the prototype.

25–50'

★LIVE BUILD

say

›Announce the flow chosen from the audience (sent during the week-gap).
›Read trigger / steps / output / metric out loud.
›Ask the author to confirm we can build it (if present).
›Build live: Zapier trigger → OpenAI classify → filter → Airtable/Sheets row → WhatsApp reply → end-to-end test.
›Every 10 min, prompt chat: "What would you change?" Read 1–2.
›If something breaks, show the debug LIVE. That's the lesson.
›Close: if you built this from zero today, it took ~15–20 real minutes. That's the point.

›Slides 15–18 (build map / step indicators).
›Share screen with creds pre-loaded.
›PLAN B running in another tab: pre-built version of the same flow. If Zapier or APIs fail, switch tabs and demo from there.
›If everything collapses: walk through one real Amalgama prototype already in production.

LIVE BUILD

Live build with chat prompts every 10'.

Don't let the build run overtime — 25 min cap, even if not perfect.

50–55'

When to scale

say

›Prototype runs out of room when: > 100 executions/day, tools-can't-handle-logic, sensitive data (PHI/PCI), cross-functional 3+ areas.
›If your business grows and a Zap isn't enough — not a problem, a signal. You validated.
›When to leave this method and level up: 2–3 quick wins live, cross-flow patterns visible, you want a 90/180/365 roadmap.

›Slides 19–21.

55–60'

CTA + close

say

›Implement your first quick win this week.
›Wellness operators: deeper conversation specifically for you — AI Blueprint Webinar A.
›Not wellness? Stay in the loop — recording + worksheet. Email opt-in.
›Thank you. Break things this week.

›Slide 22 (W-A application + QR). DO NOT mention $4,900 — that's W-B.

Q&A

Residual Q&A (3–5 min).

Strategic questions only. Sprint-specific questions → "Apply to W-A, we cover it there."

after the series

Post-webinar sequence

T+1 day: thank-you email + S2 recording + apply to W-A link.
T+7 days: nudge to wellness ops who didn't apply yet.
T+14 days: final CTA / nurture sequence for non-wellness attendees.
T+30 days: review attendance → W-A application conversion. Adjust mid-funnel messaging.