AI Commerce Creator — Slide Deck v2

OPENING: What's Possible (Slides 1–8)

Slide 1: Title

🤖💰 AI-POWERED COMMERCE CREATOR
Make Content That Sells

[Your Name / Brand]

Slide 2: Today's Promise

By 5pm today, you will have:

✅ A 15-sec polished product ad
✅ A 30-sec organic UGC review
✅ 3 social media image posts
✅ Your own AI voice + music
✅ A ComfyUI automation workflow

No design skills. No video experience. No ideation paralysis.
Just follow the cases. Build as you learn.

Slide 3: This Is Happening Right Now

[6 viral content examples — one per slide transition]

🧴 Perfume bottle, liquid swirling in slo-mo, light refracting
☕ Coffee pour — same shot, 3 different moods (Higgsfield)
🎧 Earbuds floating in space, sound waves visible
📱 "I tried this skincare for 30 days" — UGC-style
🏠 Room transforms from morning to night in one shot
💄 Lipstick color changes with camera angle shift

Slide 4: The Tools Making This Possible

TEXT    → ChatGPT, Claude
IMAGES  → DALL·E 3, Bing Image Creator
VIDEO   → Kling AI, Runway, Higgsfield Cinema Studio
AUDIO   → ElevenLabs, Suno
ASSEMBLY → CapCut, DaVinci Resolve
AUTOMATION → ComfyUI

All have free tiers. You'll use every one today.

Slide 5: Higgsfield Cinema Studio — The Camera Revolution

50+ cinematic camera presets:
  📹 Zoom    📹 Pan    🚁 Drone    🤚 Handheld    🎥 Dolly

Mood shift — same shot, different atmosphere:
  🌅 Golden Hour    🌙 Neon Night    ☁️ Moody Morning

Viral presets library:
  → Encodes what already works on social
  → One-click: "Make this product cinematic"

Single photo → full video with camera movement

Slide 6: Today's Approach

❌ No "think of an idea" (paralysis kills workshops)
✅ Pre-made cases — you just follow and build
✅ Competition — reverse-engineer an image, closest wins
✅ Real products, real formats, real platforms

You're not "learning AI."
You're making content that could sell TODAY.

Slide 7: The Flow

TEXT (copy + script)
  ↓
IMAGE (social post + UGC + storyboard)  ← COMPETITION
  ↓
VIDEO (text-to-video, image-to-video, start/end frame)
  ↓
AUDIO (voiceover + music)
  ↓
ASSEMBLY (polished ad + organic UGC)
  ↓
AUTOMATION (ComfyUI — do it all in one click)

Slide 8: Ground Rules

1. Follow the pre-made cases — no ideation required
2. Ask questions anytime — this is a workshop, not a lecture
3. Share your screen if stuck — we debug together
4. Competition is friendly — the real win is learning
5. Export everything — you'll leave with a portfolio piece

MODULE 1: AI Text Generation (Slides 9–12)

Slide 9: Module 1 Title

MODULE 1
AI Text Generation

✍️ Captions, Ads, Scripts

Slide 10: Your Case

PRODUCT: HydraGlow Smart Water Bottle
PRICE: $49
FEATURES: Tracks hydration, glows to remind you,
          keeps water cold 24 hours
AUDIENCE: Health-conscious millennials

YOU NEED:
1. Instagram caption (hook + body + CTA)
2. Ad headline + description (for paid ads)
3. 15-second video script (shot-by-shot + VO)

Slide 11: The Formulas

AIDA: Attention → Interest → Desire → Action
  "You're dehydrated right now and don't know it." → 
  "This bottle tracks every sip." → 
  "Imagine never forgetting to drink water again." →
  "Tap the link. Your body will thank you."

PAS: Problem → Agitation → Solution
  "80% of adults are chronically dehydrated." →
  "It's killing your energy, skin, and focus." →
  "HydraGlow. The bottle that won't let you forget."

HOOK-FIRST: You have 1.5 seconds on social.
  Open with: Question. Statistic. Bold claim. Controversy.

Slide 12: EXERCISE — Module 1

1. Open ChatGPT
2. Use the pre-made prompt (next slide)
3. Generate: 1 caption, 1 ad, 1 video script
4. Read your best hook aloud to the group

PRE-MADE PROMPT:
"You're a direct-response copywriter. Write 3 pieces for HydraGlow,
a $49 smart water bottle that tracks hydration, glows to remind you
to drink, and keeps water cold 24 hours. Target: health-conscious
millennials. Deliver: (1) Instagram caption with hook, (2) Facebook
ad headline + 2-line description, (3) 15-sec video script shot-by-shot
with voiceover script."

MODULE 2: AI Image Creation (Slides 13–20)

Slide 13: Module 2 Title

MODULE 2
AI Image Creation

📸 Social Posts, UGC Photos, Storyboards

Slide 14: Prompt Anatomy

A good image prompt has 5 parts:

1. SUBJECT — What's in the frame?
   "Cold brew bottle, marble surface, coffee beans"

2. STYLE — What kind of image?
   "Flat lay product photography" / "Casual candid photo"

3. LIGHTING — How is it lit?
   "Warm morning light through window" / "Studio softbox"

4. COMPOSITION — Camera framing
   "1:1 square" / "16:9 cinematic" / "Overhead shot"

5. QUALITY TAGS — Technical descriptors
   "Photorealistic, 4K, shallow depth of field"

Slide 15: Your Case

BRAND: Mornings Coffee
PRODUCTS: Cold brew, single-origin beans
AESTHETIC: Minimalist, warm, design-conscious
AUDIENCE: Coffee lovers, 25–40

YOU NEED:
1. Social media post (Instagram carousel-worthy)
2. UGC-style photo ("real person" holding product)
3. Storyboard frame (cinematic, for video ad)

Slide 16: The 3 Image Types

SOCIAL POST:
"Flat lay product photography, Mornings cold brew bottle
on marble surface, coffee beans scattered, warm morning
light, minimalist, 1:1 square, Instagram aesthetic"

UGC PHOTO:
"Casual photo of person holding Mornings coffee bottle,
sitting at kitchen table, morning sunlight through window,
candid, unposed, phone camera quality, natural skin texture"

STORYBOARD:
"Cinematic wide shot, cold brew pouring into glass with
ice, condensation on bottle, golden hour backlight,
shallow depth of field, 16:9, commercial quality"

Slide 17: COMPETITION — Reverse Prompt Engineering

THE CHALLENGE:
I'll show you a reference image.
You write a prompt to recreate it.
Closest match wins.

3 rounds. Increasing difficulty.

WHY THIS WORKS:
→ Forces precision (every word counts)
→ Teaches prompt anatomy better than any lecture
→ No "I don't know what to make" — the target is right there
→ Competition = memory = retention

Slide 18: Competition — Round 1

[Show reference image: Simple product on white background]

ROUND 1: Easy
A single product. Clean background. Studio lighting.

What to nail:
- The product itself (color, shape, details)
- The background (color, texture)
- The lighting direction

Write your prompt. Submit via [shared doc / chat].

Slide 19: Competition — Round 2

[Show reference image: Product in lifestyle setting]

ROUND 2: Medium
Product in a scene. Specific lighting. Props.

What to nail:
- Everything from Round 1, PLUS
- The setting (room type, surfaces, time of day)
- The lighting quality (golden hour, window light, shadows)
- The composition (angle, framing, depth of field)

Write your prompt. Submit.

Slide 20: Competition — Round 3

[Show reference image: UGC-style with "imperfections"]

ROUND 3: Hard
"Real person" holding product. Unpolished look.

What to nail:
- Everything from Round 2, PLUS
- "Phone camera" aesthetic (not professional)
- Natural skin texture (not airbrushed)
- Candid framing (not posed)
- Imperfections (slight blur, natural light, unposed)

Write your prompt. Submit.

MODULE 3: AI Video Generation (Slides 21–27)

Slide 21: Module 3 Title

MODULE 3
AI Video Generation

🎥 Text → Image → Video

Slide 22: Three Ways to Generate Video

1. TEXT-TO-VIDEO
   Prompt only → AI creates everything
   Fastest. Least control.

2. IMAGE-TO-VIDEO
   Upload keyframe → AI animates it
   Most control. Needs good starting image.

3. START + END FRAME
   Upload image A AND image B → AI fills the transition
   Best for before/after, transformation, reveals.

Slide 23: Your Case

PRODUCT: PulsePods Wireless Earbuds
PRICE: $129
FEATURES: 36hr battery, water resistant, spatial audio,
          active noise cancellation

YOU NEED:
1. Text-to-video clip (abstract product shot)
2. Image-to-video clip (UGC person using product)
3. Start+end frame clip (case → in-ear transition)

Slide 24: Clip 1 — Text-to-Video

TOOL: Kling AI or Runway
INPUT: Text prompt only

"Cinematic close-up of wireless earbuds floating in
space, pulsing sound waves visible as blue light,
dark background with subtle particles. Smooth rotation.
5 seconds."

WHY: Abstract product shots are hard to photograph
but AI generates them perfectly. This is the "hero shot."

Slide 25: Clip 2 — Image-to-Video

TOOL: Kling AI or Runway
INPUT: Your UGC photo from Module 2 + prompt

"Person puts earbuds in, expression shifts to delight
as music starts, natural head nod, candid moment.
Handheld camera feel. 5 seconds."

WHY: The UGC photo sets the character. The prompt
animates the reaction. Together = authentic demo.

Slide 26: Clip 3 — Start + End Frame

TOOL: Kling AI or Runway
INPUT: Image A (case closed) + Image B (in ears) + prompt

"Earbud case opens smoothly, earbuds float upward,
rotate and slide into ears. Smooth tech aesthetic.
Clean transition. 5 seconds."

WHY: Transformation is the most engaging video format.
Viewers stay to see the "after." Perfect for product demos.

Slide 27: Higgsfield — Camera Without Complexity

SAME PRODUCT. 9 DIFFERENT LOOKS.

Camera presets (pick from menu):
  🔍 Slow zoom in    ↔️ Pan across    🚁 Drone rise

Mood presets (click to apply):
  🌅 Golden Hour    🌙 Neon Night    ☁️ Moody Morning

Single photo → cinematic video. One click.

[Show demo: same product image, 3 camera presets × 3 moods]

MODULE 4: AI Audio (Slides 28–31)

Slide 28: Module 4 Title

MODULE 4
AI Audio Generation

🗣️ Voiceover + 🎵 Music = Brand

Slide 29: The Two-Brand Test

SAME COPY. DIFFERENT VOICE. DIFFERENT BRAND.

Product A: LuxeGlow Serum ($89, premium skincare)
  Voice: [Warm, sophisticated, intimate]
  Pace: Slow, deliberate
  Music: "Luxury spa ambient, soft piano, gentle strings, 70bpm"

Product B: BeatBuds ($49, budget tech)
  Voice: [Energetic, excited, casual]
  Pace: Fast, punchy
  Music: "Upbeat tech pop, punchy drums, synth bass, 120bpm"

Voice IS your brand signal.

Slide 30: EXERCISE — Voice Cloning

1. Record 60 seconds of your voice (phone, quiet room)
2. ElevenLabs → VoiceLab → Add Voice → Instant Clone
3. Upload recording → name your voice
4. Generate LuxeGlow VO:
   "[Warm, sophisticated] Your skin deserves more than hope."
5. Generate BeatBuds VO:
   "[Energetic, excited] 36 hours of battery. Zero excuses."

LISTEN: Same person's voice. Two completely different brands.
That's what emotion tags do.

Slide 31: EXERCISE — Music

SUNO — 2 tracks, 2 vibes:

Track 1 (LuxeGlow skincare):
"Luxury spa ambient, soft piano, gentle strings,
70bpm, C major, no drums, warm atmosphere"

Track 2 (BeatBuds tech):
"Upbeat tech pop, punchy drums, synth bass,
120bpm, E minor, energetic, short"

MUSIC = MOOD.
Match the music to the product, not your personal taste.

MODULE 5: Commercial & UGC (Slides 32–36)

Slide 32: Module 5 Title

MODULE 5
Commercial & UGC Assembly

💰 Sell the product. Two ways.

Slide 33: Two Formats, One Product

POLISHED AD (15 sec)              ORGANIC UGC (30 sec)
─────────────────────              ────────────────────
Cinematic, color-graded            Phone-selfie, natural
Professional VO                    Conversational VO
Brand-matched music                Trending/lo-fi music
Key features + price               Personal testimony
"Available now. Link in bio."      "I paid for these. Honest
                                   review. Link below."

YOU NEED BOTH.
Polished = credibility. UGC = trust.

Slide 34: Assembly — Polished Ad

TIMELINE (CapCut or Resolve):

0:00-0:03   Clip 1 (text-to-video abstract shot)
            VO: "36 hours of battery."
0:03-0:06   Clip 3 (start→end frame transition)
            VO: "Noise cancellation that actually works."
0:06-0:09   Clip 2 (person using product)
            VO: "PulsePods. Hear what matters."
0:09-0:12   Product shot + text overlay: "$129"
0:12-0:15   Logo + "Link in bio."

Music: BeatBuds tech track. Duck during VO.

Slide 35: Assembly — Organic UGC

TIMELINE (CapCut or Resolve):

0:00-0:05   "Phone selfie" intro — person talking to camera
            VO: "Okay so I just got the PulsePods..."
0:05-0:10   Clip 2 (person using, natural reaction)
            VO: "...and honestly the battery is insane."
0:10-0:18   Clip 1 (abstract, overlaid with casual commentary)
            VO: "I've been using them for a week. No issues."
0:18-0:25   Clip 3 (transition, product close)
            VO: "Not sponsored. I paid $129. Link below."
0:25-0:30   Text: "Honest review ↓" + CTA

Music: Trending lo-fi or no music (natural feel).

Slide 36: Affiliate CTAs That Convert

WHERE TO PUT IT:
  → Last 2-3 seconds of video
  → Peak attention (right after the reveal or testimonial)

WHAT TO SAY:
  "Link in bio 👆"
  "Comment 'LINK' and I'll DM you"
  "Tap the link to shop"
  "Check the description"

WHAT NOT TO SAY:
  ❌ "Buy now" (too aggressive)
  ❌ "Limited time" (unless true)
  ❌ CTA at the START (nobody's ready yet)

MODULE 6: ComfyUI (Slides 37–41)

Slide 37: Module 6 Title

MODULE 6
ComfyUI Workflow Builder

⚡ Automate everything you did today

Slide 38: Why Automation Matters

What you did today (manual):
  Prompt → generate image → download → upload to video tool
  → generate video → download → upload to editor → repeat

What ComfyUI does:
  Prompt → [Workflow loads model → generates image
  → passes to video node → generates video → saves]
  ONE CLICK.

50 product images/week, manually:
  5 hours.

50 product images/week, ComfyUI:
  10 minutes.

Slide 39: The Node Paradigm

[Show ComfyUI interface screenshot]

Each box = one AI operation.
Connect with wires = define the flow.

SIMPLE WORKFLOW:
[Load Model] → [Type Prompt] → [Generate Image] → [Save]

PRODUCT AUTOMATION:
[Load Model] → [Type Prompt] → [Generate Image]
  → [AnimateDiff] → [Generate Video] → [Save Video]

Build once. Run 1000 times.

Slide 40: EXERCISE — Build Your First Workflow

1. Open ComfyUI (demo station or your GPU laptop)
2. Load pre-built workflow: product-image.json
3. Change the prompt to YOUR product
4. Click "Queue Prompt"
5. Watch the pipeline execute

You just automated product image creation.

Next: Load product-video.json
→ Same concept, adds AnimateDiff node
→ Static image → animated video in one click

Slide 41: Where This Leads

What professionals are building:

→ Batch systems: 100 product images from CSV of product names
→ Template pipelines: Change 1 prompt → full video ad updates
→ Multi-platform: Same workflow, different aspect ratios
→ API integration: Connect to Shopify, schedule posts

Today: You learned to make 1 ad.
ComfyUI: You learned to make 1000.

Figma Weave = cloud/polished version of same concept.
But ComfyUI is free, local, unlimited. Start here.

CLOSING (Slides 42–44)

Slide 42: What You Built Today

In 8 hours, you:

✍️ Wrote copy for a real product
📸 Created social posts + UGC photos + storyboards
🎥 Generated video 3 different ways
🗣️ Cloned your voice + scored your ad
💰 Assembled a polished ad AND an organic review
⚡ Built an automation workflow in ComfyUI

This portfolio didn't exist this morning.

Slide 43: Keep Building

Tools to master next:
  ComfyUI    → Deep automation, custom pipelines
  Higgsfield → Camera presets, viral templates
  ElevenLabs → Voice design, multi-voice projects
  Runway     → Professional video generation
  Kling      → Longer, more complex video

Concepts to explore:
  → Start/end frame narratives (transformation sells)
  → UGC authenticity (imperfection = trust)
  → Prompt precision (the competition proved this)
  → Affiliate content systems (build once, sell repeatedly)

Slide 44: Thank You

💰🤖

The tools are ready.
The cases are real.
Go make content that sells.

[Contact info]
[Link to all workshop docs + cheat sheet]

BONUS: Lip Sync (Slides B1–B2)

Slide B1: When Commerce Needs Lip Sync

👄 + 🔊 = 🎬

Only sync lips when:
  ✅ Lips are clearly visible
  ✅ Shot is close-up
  ✅ Voice is meant to come from that person

Skip lip sync when:
  ❌ Voiceover over product shots
  ❌ Wide shots (mouths invisible)
  ❌ Cutaways, B-roll, text-only

TOOLS:
  Sync.so → Upload video + audio, get synced result
  HeyGen → Talking-head avatar, type text, it speaks

Slide B2: Demo

1. Take your UGC video (person using product)
2. Take your VO from Module 4
3. Upload both to Sync.so
4. Download synced result
5. Compare: unsynced voiceover vs lip-synced

The difference matters for talking-head content.
For product B-roll? Nobody notices.

Deck Production Notes

Media to Prepare

6 viral content screenshots (X/Instagram)
Higgsfield demo video (1 product, 3 camera angles × 3 moods)
3 competition reference images (easy, medium, hard)
ComfyUI screenshot (simple workflow)
CapCut/Resolve timeline screenshot (polished ad)
Before/after: unsynced vs synced (bonus module)
Final "Aurora" or "PulsePods" demo video

Speaker Notes

Transitions between modules
Competition rules (verbatim, on screen)
Timing warnings (visible only to presenter)
FAQ answers (see appendix)

Appendix: FAQ

Q: Will AI replace content creators? A: It replaces the repetitive parts of creation — generating options, resizing formats, trying variations. It doesn't replace taste, strategy, or knowing what will resonate with an audience. The creators who thrive are the ones who learn these tools.

Q: Can I use AI-generated content for client work? A: Yes — with paid tiers. Free tiers are for learning. Check each tool's commercial terms. Voice cloning requires consent.

Q: How realistic is UGC-style AI content? A: Very — if you prompt for imperfections. The key is asking for "phone camera quality," "natural skin texture," and "candid, unposed." Perfect AI output looks fake. Imperfect AI output looks real.

Q: Why ComfyUI and not just keep using Kling/Runway? A: Per-use costs add up at scale. ComfyUI = unlimited generation on your own GPU. For professional content creators producing volume, it pays for itself almost immediately.

Q: What's the most important skill to develop after this workshop? A: Taste. The AI generates. You select. The difference between good content and great content is the human choosing which of the 10 AI-generated options is actually worth publishing.