正在加载视频...

视频加载失败

I built an AI system that automates product video creation for entire e-commerce catalogs. (Saves $30K per collection shoot, increases conversion rates by 40% on sites) Here's how the automation works: → Firecrawl scrapes product photos from any e-commerce product page → Nano Banana create 4 images showcasing each...

170,695 次观看 • 9 个月前 •via X (Twitter)

37 条评论

David Roberts 的头像
David Roberts9 个月前

Full step-by-step setup video:

PK @ Realtime 的头像
PK @ Realtime9 个月前

I would hate to see the API bill but like your combo with firecrawl 🔥

David Roberts 的头像
David Roberts9 个月前

😅

Amrit Chera 的头像
Amrit Chera9 个月前

It's risky to do this because consumers really care about the finer details, and AI often misses those. Even minor color differences make a big difference. This is only good in theory

Bugante 的头像
Bugante9 个月前

The conversion lift is real but raw Veo outputs will tank your Shopify page speed. You need to add a compression node to get files under 2MB. If the page loads slow the bounce rates eat the conversion gains instantly. Performance matters more than aesthetics.

Tristan Rhodes 的头像
Tristan Rhodes9 个月前

1) You should allow sending multiple images to Veo so it can see multiple angles of the products (including the backside) for accuracy 2) You could consider asking Nano Banana to put views from all 4 angles in a single image (2x2 grid). This will save costs. Cool project and video!

Rea 的头像
Rea9 个月前

MOTION

AILegend 的头像
AILegend9 个月前

Motion

Dirck 的头像
Dirck9 个月前

MOTION

fredette 的头像
fredette9 个月前

Motion

Nati Adler 的头像
Nati Adler9 个月前

MOTION

Anayat 的头像
Anayat9 个月前

The Firecrawl Nano Banana Veo 3.1 chain is cool but the real unlock is treating each step as a typed tool with usage metrics so you can auto swap models, cap API burn, and reuse the workflow across brands instead of rebuilding per client.

Armando Pérez Cadena 的头像
Armando Pérez Cadena9 个月前

MOTION

Siméone 的头像
Siméone9 个月前

MOTION

Matt Caron 的头像
Matt Caron9 个月前

Motion

MISFITSッ 的头像
MISFITSッ9 个月前

motion

Tony Crisp 的头像
Tony Crisp9 个月前

Motion

James Edison 的头像
James Edison9 个月前

Nice build. Only thing I’d stress: “+40% conversion” needs clean attribution. Most wins here come from higher PDP engagement + lower returns, not magic. If you can show: - baseline vs test (same traffic source) - lift in ATC, not just views - where it breaks (bad lighting, reflective products, variants) …this becomes a serious DFY offer for DTC, not just a cool workflow.

Digital_Observer 的头像
Digital_Observer9 个月前

MOTION p.s. I would change Firecrawl with @apify 😇

giobeau 的头像
giobeau9 个月前

MOTION

Terra’s Dad 🪐 的头像
Terra’s Dad 🪐9 个月前

Motion

daan 的头像
daan9 个月前

MOTION

Juan Pablo Murcia 的头像
Juan Pablo Murcia9 个月前

Motion

Stone May 的头像
Stone May9 个月前

motion

Luke Cody 的头像
Luke Cody9 个月前

Motion

Social Monkey 的头像
Social Monkey9 个月前

The combination of scraping product photos and automating video creation is smart. I bet the next step is personalizing those videos based on user data.

Jay Singh Kang 的头像
Jay Singh Kang9 个月前

MOTION

David Roberts 的头像
David Roberts9 个月前

gotcha! will send over

luis carlos cepeda giraldo 的头像
luis carlos cepeda giraldo9 个月前

Motion

Kim🌊 的头像
Kim🌊9 个月前

this is gonna absolutely cook for niche DTC brands

John Victor 的头像
John Victor9 个月前

Motion

Ahmet Kabaci 的头像
Ahmet Kabaci9 个月前

MOTION

when to moon 的头像
when to moon9 个月前

MOTION

Kritarth Mittal | Soshals 的头像
Kritarth Mittal | Soshals9 个月前

Let’s talk! I’m working on a v cool project, might interest you

🇬🇧 UKMiller (XRP 💎 👐) 的头像
🇬🇧 UKMiller (XRP 💎 👐)9 个月前

MOTION

elmoustache 🧔 的头像
elmoustache 🧔9 个月前

MOTION

TheDropPirate 的头像
TheDropPirate9 个月前

MOTION

相关视频

Endless influencer ads without shipping/mailing a single product. (How I automated e-commerce creative generation and cut campaign costs by 75%) Traditional influencer marketing sucks for most e-commerce brands: → High upfront costs for product samples → Shipping delays that kill campaign momentum → Limited creative output from each collaboration → No way to test concepts before major investment Google's Nano Banana changes all of that. I built this n8n automation that generates unlimited influencer ad creative using just product photos and a reference image of your influencer: → AI combines your product with influencer photos seamlessly → Creates multiple campaign variations in minutes → Generates high-quality, campaign-ready assets → Eliminates sample costs and shipping delays → Perfect for A/B testing creative concepts The business impact has been massive: ✅ 75% reduction in campaign setup costs ✅ 10x faster creative iteration cycles ✅ 300% increase in testable ad variations ✅ Zero inventory risk for new product launches This system reclaimed 15+ hours per week that used to go toward managing physical product samples and shipping logistics. The complete automation runs on n8n with Google's Imagen 3 API – no coding required. Want to see exactly how it works? 1. Like & RT 2. Bookmark this post for later reference 3. Comment "SYSTEM" below (must be following so I can DM you) I'll send you the entire system template for free, including setup instructions and prompt engineering guides.

David Roberts

26,261 次观看 • 11 个月前

Every e-commerce store needs product photos… so I automated the entire process with AI + n8n. This end-to-end workflow turns one iPhone photo into 8–10 professional product shots, handling everything from model generation to final delivery. I just dropped a detailed breakdown covering: → How AI analyzes products and generates the right models → How meta-prompting creates dynamic scenes for each item → How the system maintains visual consistency across every image → How to customize this for any product category The real magic is the prompting strategy. Instead of static prompts, the system analyzes the uploaded product and generates contextually perfect scenes. Upload athletic wear → gyms and outdoor running shots Upload formal wear → office and event settings Upload winter gear → mountain lodges and snowy landscapes This has already helped e-commerce brands cut photography costs by up to 90% while drastically speeding up launches. Just think… there are 30M+ e-commerce stores, and all of them need high-quality product photos. I’m not saying this replaces professional photography entirely—but for product testing, rapid launches, and catalog builds, it’s a game-changer. (Especially when you’re generating variants while competitors are still booking shoots.) Want the exact n8n template + prompts behind this? Like & RT Follow me (so I can dm you) Comment “CATALOG” I’ll send the full n8n JSON for free + access to the YouTube walkthrough showing the full build.

David Roberts

15,391 次观看 • 8 个月前

Every e-commerce store needs product photos... so I automated the entire process with AI and n8n. This end-to-end workflow transforms a single iPhone photo into 8-10 professional product shots, handling everything from model generation to final delivery. I just dropped a detailed breakdown covering: → How the AI analyzes products and generates perfect models → How meta-prompting creates dynamic scenes for the model and article of clothing → How the system maintains consistency across all generated images → How you can customize this for any product category The magic happens in the prompting strategy. Instead of writing static prompts, I use AI to analyze the uploaded product and generate contextually perfect scenes. Upload athletic wear? It creates gym and outdoor running scenarios. Upload formal wear? It generates office and event settings. Upload winter gear? Mountain lodges and snowy landscapes. This system has already helped several e-commerce clients slash their photography costs by 90% while dramatically reducing time-to-market. Just think... there are over 30 million e-commerce stores out there, and they ALL need high quality product photos!!! I'm not saying this will replace all professional photography, but for product testing, rapid launches, and catalog building, it's a game-changer. (Especially when you can generate variations while your competitors are still booking photoshoots) Want the exact n8n template and prompts powering this system? 1. Like & RT this post 2. Follow me for more AI automation breakdowns 3. Comment "CATALOG" below I'll send you the complete n8n JSON for free, plus access to my YouTube tutorial showing the step-by-step build.

David Roberts

89,185 次观看 • 11 个月前

While everyone's waiting 5 years for Sora to generate one extremely woke and mediocre image, I built an agent that creates unlimited product variations in minutes. This digital sweatshop takes your product and drops it into any reference scene you want - no photographers, no studio rentals, no dealing with "creative differences." Perfect for anyone who needs 100+ product images but don't want to remortgage their house paying agencies. Here's what this n8n workflow does: → Pulls reference images from your Google Drive folder → Grabs all your product photos from another folder → Creates every possible combination automatically → Generates professional product scenes using AI → Saves all variations back to Drive, organised and ready Instead of spending $500 per product shot, you spend $0.03 per image and get infinite variations. While your competitors are: → Booking photographers for $2K shoots → Waiting weeks for "creative concepts" → Getting 10 images they hate → Dealing with revision cycles from hell You're getting: → Hundreds of product variations instantly → Consistent brand aesthetic across all images → Zero human drama or creative ego trips → Images that actually convert instead of "looking artistic" This thing runs 24/7. Feed it new reference scenes, and it automatically generates fresh product images while you sleep. E-commerce brands are about to realise they've been massively overpaying for something a weekend workflow can handle. Follow + RT + Comment "PRODUCT" and I'll send the JSON file that turns your Google Drive into an automated creative department. (Need to do all 3 or my agent won't find your handle to send the goods)

Nozz

51,321 次观看 • 1 年前

Nano Banana Pro + n8n is absolutely insane 🤯 Google just released Nano Banana Pro and I already have it running inside n8n + Airtable. The use case? Steal your competitors' Facebook static ads. Perfect for DTC brands & agencies who see competitor ads they want to replicate with their own products. Nano Banana Pro is SICK: → 4K resolution output (studio quality) → Multiple image inputs (up to 14 images in one generation) → Perfect text rendering in multiple languages → Maintains consistency across up to 5 people Here's the workflow: You see a killer competitor Facebook ad. Perfect layout, colors, messaging—but it's their product, not yours. Instead of briefing a designer and waiting days for recreations... This automation does it in 30 seconds: → Upload competitor ad image in Airtable → Upload your product image → Click "Generate" button → Nano Banana Pro swaps ONLY the product → Keeps everything else identical (layout, fonts, colors, graphics) → Returns 4 variations automatically → All delivered back to Airtable instantly No designer bottlenecks. No recreation from scratch. No guessing what works. What you get: → Competitor-inspired ads with your product → Same proven layout and design language → 4 variations per request for testing → Everything organized in Airtable → Feels like a mini creative co-pilot Built 100% in n8n + Airtable. I filmed a 30-minute Loom video walking you through how to set it up step-by-step. Want the n8n template, Airtable base, and Loom video for free? > Comment "BANANA" > Like this post And I'll send it over (must be following so I can DM)

Mike Futia

88,743 次观看 • 10 个月前