Former Liquid Death & Tastemade Marketing Strategist | MBA | Boosting socials, steering through Web3, and befriending AI 🤖

I build brands, grow engaged communities, and drive real impact. With 20+ years in digital marketing, I’ve helped companies scale from zero to millions in reach, leading campaigns that get attention and convert.

At Hexens, I drove marketing for the largest Web3 security competition ever. At BitcoinOS, I built a 30K+ member community in just three months. At Liquid Death, I led viral campaigns, generating 3B+ impressions. I’ve also managed 3,000+ influencer partnerships and scaled Web3 and gaming communities to 500K+ members.

Whether it’s organic growth, paid media, influencer marketing, or community strategy, I create high-impact, data-driven marketing that delivers results. Let’s build something that lasts.

Shahab Shirmohammadi

Optimizing Fashion Industry processes with 3D & AI | Whispering the Secrets of Fashion Tech | Keynote Speaker & Fashion Geek

I started as a seamstress 15 years ago because I couldn't afford fashion school.
Now I'm building AI that makes actual garments, not Instagram content.

Long story short: 
2009: Lost everything in my first fashion business.
2012: Borrowed €5K for Sri Lanka production. Sold 3 garments. Failed miserably.

Rock bottom taught me what success never could.
I begged for work on Alibaba. Got gigs. Saw factory floors worldwide. Learned pattern making doing it 10 hours daily at Max Mara. Then Timberland.

Spent years watching the same stupid inefficiencies:
→ Pattern makers recreating patterns that already exist
→ Designers making pretty pictures nobody can produce
→ Companies spending €450K yearly on patterns, €90K duplicates

2023: Tried GenAI. August: garbage. November: holy sh*t.
The speed of change was insane.
2024: Couldn't find a sketch-to-pattern tool that actually worked.

LinkedIn post appeared - Liquid Factory competition. €200K for 4 founders.
I was 37, steady corporate money.
My dream was stronger than my fear.

Got selected from 181 applications. 
Left Timberland. 
Started fashionINSTA.

What I actually built:
I don't make pretty AI images. I make patterns factories can cut and sew.
My clients discovered they're recreating 20% of patterns they already own.
One was wasting €90K yearly on duplicate work.

The pivot that mattered:
Beta testers: "Cool, but I'd do it differently."
Every pattern maker said this. There's no "right" way.
I threw out our code. Started from scratch.
Built AI that learns YOUR way, not generic templates.
We train on YOUR patterns. Your armholes. Your fit.
750 data points per pattern. Pattern graveyard becomes goldmine.

Results:
Pattern search: 4 hours → 2 seconds
New patterns: 5 days → 15 minutes
Development: 70% faster

Technology should amplify expertise, not replace it.

I'm building the first NO BS fashion AI tool. 
Because fashion needs production, not pixels.

Currently serving enterprise clients tired of AI fantasy clothes.

Mission: Make pattern making fun again.

Want real AI pattern making? Let's talk.

Sylwia Szymczyk

For the past 15 years, I've been on a mission to make technical knowledge matter in fashion.
 
We have a big problem in this industry, and few people recognize it.
 
30% of fashion graduates never used a sewing machine. 
Pattern makers retiring faster than we're training new ones. 
Designers who don't know how garments are made.
 
The fact is, we're losing technical literacy.
 
There is no AI tool. 
There is no 3D software. 
There is no prompt.
 
That can replace understanding how real fabric behaves on a body.
 
Meanwhile, everyone's chasing photorealistic renders and viral AI images.
 
And yet... pattern making is the foundation of every garment ever sold.
 
It's more important than ever to bridge the gap between creativity and construction.
 
It's up to us to pass this knowledge forward before it disappears.
 
P.S. Tag someone who still believes technical skills matter below!

NO BS fashion AI | Seamstress @Max Mara → Innovation lead @Timberland → fashionINSTA founder

What if you could change your pattern and see the AI garment update in real time?
 
No manual rendering that takes forever.
No adjustments in the 3D window.
No need for a fast PC.
No re-simulation.

Just change the geometry. 
See the result. Instantly.
 
This is what we're building at fashionINSTA.
 
Make a cut. See the garment. 
Add gathering. See the garment. 
Adjust a dart. See the garment.
 
AI preview is directly connected to your.DXF geometry. 
Changes apply immediately.
 
Now think bigger.
 
Different colourways.
Random color blocking.
Hundreds of cut variations.
Small adjustments across different fabrics.
 
All at scale. 

All connected to geometry that can actually be cut and stitched.
 
Not renders you'll need to recreate. 
Real patterns. Real previews. Real production.
 
This is how design and PD workflows should work.
 
And this week's update makes it real.
 
Stay tuned. Next weeks will bring even hotter ones!

The most dangerous assets in a fashion brand today are photorealistic AI visuals.
 
Not because it looks bad. 
Because it looks too good.
 
And it's breaking your company in 3 ways.
 
1. Choice overload is killing decisions.
 
Designers generate thousands of visuals in seconds. 
But no criterion to choose the right ones.
More images. Worse decisions.
 
2. Internal sales got easier. Delivery got harder.
 
AI makes it easy to sell ideas internally. 
Show merch a stunning visual. 
Get leadership excited.
 
The trick? 

Everyone believes this will be the exact garment they'll get.

Then development starts. 
Garments dropped or are completely different. 
Disappointment all around.
 
3. Your silos just got tighter.
 
Merch thinks designers aren't needed anymore. 
"Anyone can type a prompt."
 
PDs receive impossible garments. 
More dropped styles. Longer back and forth.
 
Designers see PD as an even bigger bottleneck.
 
Everyone's frustrated. Nobody's talking.
 
 
This is why we built fashionINSTA differently.
 
Generate an idea. 
Instantly know if it's feasible. 
Instantly estimate the cost.
 
Out of 1000 AI images, maybe 10% can actually be made. 
We tell you which ones.
 
Our visuals aren't fantasies. 
They're connected to real.DXF patterns.
What you sell internally is what you can produce.
No surprises.
 
And designers don't just hand over a visual. 
They deliver a first pattern draft that preserves brand identity. 
Saving pattern makers 50% of effort.
 
That's how you stop being enemies. 
That's how you become buddies.
 
AI visuals aren't the problem. 
Disconnected AI visuals are.

We've done it again.
 
True .DXF → AI image connection in action.
 
1. Place your AOP
2. Adjust dimensions
3. See it in AI preview
 
Instantly.
 
Still fine-tuning, but going live next week.
 
True color blocking is driven by the actual print. 
No masking without real-life reference. 
No guessing dimensions or placement.
 
The precision fashion actually needs.
What you see is what you can cut.
 
Get on the waitlist to try it in our node-based version.
 
https://lnkd.in/dJ8Me6AU

Product development is easy.
 
All you have to do is...
 
→ Make the sample match the impossible render
→ Translate "make it elevated" into specs
→ Manage 47 styles across 12 factories
→ Explain to design why physics exists
→ Hit target price with uncosted fabric
 
Simple.
 
People think we just "send emails" and "check samples."
 
INSTEAD
 
We take heat and fix fires silently.
We negotiate reality with dreamers.
We protect margin while protecting vision.
We bridge imagination and the factory floor.
We hold 200 details, while others hold opinions.
 
Some days, smooth approvals and on-time shipments. 
Other days, the button supplier ghosted us. 
And we ship in 72 hours anyway.
 
The magic of great product development?
You make chaos look like a machine.
 
Which factory disaster still makes you laugh?

PS. Images generated in fashionINSTA. With one click.

Why PD is fashion companies' coolest department?

We turn sketches into what pays everyone's salary. 

Nobody knows we exist.

Garment is sold and worn because PD made it possible. 

Not designed it. 
Not marketed it. 
Made it manufacturable.
And under required margins.

We are product developers, 
technical designers
pattern makers.

After the collection is designed:

Designer: "Architectural shoulders" 
Factory: "€15 materials alone" 
PD: Strategic darts + fusible = same look, 70% less

Merch: "12 colors, 200 units each" 
Factory: "500 minimum per color" 
PD: Switch to garment-dye at 100pc MOQ

A sketch is paper until PD touches it.

We protect design intent while ensuring survival. 
Too many compromises? Product dies. 
Too few? Business dies.

Every garment has 100 invisible decisions: 
→ Which interfacing? 
→ Bias cut without killing margins? 
→ Garment-wash vs enzyme-wash? 
→ 3 colors or 5?
Each decision makes or breaks it.

We touch everything: 
→ Design vision. 
→ Merch strategy. 
→ Finance margins. 
→ Factory limits. 

If clothes don't ship, nobody gets paid.

Fashion schools teach design. 
Nobody truly teaches PD. 
You learn by failing.

No fashion week applause. 
No Instagram credit. 

But that garment you're wearing? 
We fought for every seam.

We don't crush creativity. 
We make it survive capitalism.

Who else thinks PD deserves recognition?

Your AI tool works for 50% of your collection.
And breaks the 50% that actually matters.

You can divide your development efforts into 2 types:
Carryovers
Newness
 
Carryovers are simple.
 
→ The pattern exists
→ The cost is known
→ Same style, seasonal colors
→ New buttons, different trims or label placement
 
AI visual of a carryover? Fine. 
Nobody's getting disappointed.
 
Newness is different.
 
→ Multiple approval rounds
→ Pattern makers staring at impossible constructions
→ Leadership in love with something that can't be made
→ Fittings that reveal "this doesn't drape like the picture"
 
Carryovers don't need selling. They're proven.
 
Newness is where collections get their identity. 
Where brands stand out.
 
And where disconnected AI visuals hurt the most.
 
Why are some brands heavily relying on carryovers?
 
Your carryover/newness ratio isn't always a strategy. 
It's dictated by where your pattern makers sit.
 
→ Overseas technical team? 
You're running 70-80% carryover. Not by choice. 

Because newness costs too much when your pattern maker is 8 time zones away.
 
→ In-house sample room? 
You can push 60%+ newness. 

Idea Monday, pattern Tuesday, sample Friday.
 
Two realities. Two different problems.
 
Who needs newness:
 
Fashion week brands.
You can't show the same stuff on the catwalk every season. Carryover collections don't get editorial coverage.
 
Fast fashion. 
Surprising, right? But what's trending on TikTok needs to be in stores in 4 weeks. That's rapid iteration, not carryover.
 
That's why AI in the discovery stage can be a real game changer, if connected properly to the real, hard data about the pattern and fit.
 
AI discovery that connects to development.
 
→ Explore 50 directions in an hour 
→ Play with demand data visually 
→ Find unexpected combinations
 
Not: "Here's a pretty picture. Figure it out."
 
But: "Here's the idea. Here's the feasibility. Here's the cost. Here's a pattern draft."
 
That's when AI stops wasting your time.
 
 
PS: Which scenario is yours?
→ Sourcing overseas, high carryover? 
→ Internal team, pushing newness? 
→ Somewhere in between?

There is no mercy in the genAI space.
 
If the only thing you do is nice images, 
Google, Anthropic, and OpenAI already killed your startup.
 
We're building a node-based system for fashionINSTA. 
From idea to ready-to-manufacture.
 
Of course, a big chunk of this is images.
 
We need them: 
→ Realistic
→ Beautiful
→ Consistent
→ Driven by user intent and taste
→ Driven by colors, aesthetic, mood
 
And you can get ALL OF THAT 
using NanoBanana and Claude.
 
We're developing using only those 2 tools right now. 

Testing everything.
 
And honestly? You might not need nodes or ANY other tool if you can prompt correctly.
 
Custom AI training is no longer necessary for images. 
Give NanoBanana your logo in a prompt. 
It places it perfectly.
 
So what can save vertical genAI tools?
 
Only 3 things:
 
→ Very edge use cases where generic models can't deliver (and fashion images are not one of them)
 
→ User interface that removes the need to write complex prompts
 
→ Chains and workflows that connect multiple tools and AI with computing
 
This is why 3D + AI tools are great. 
This is why fashion designers love Weavy or Flora. 

This is how we build fashionINSTA - connecting images with geometry and hard data.
 
Pretty pictures are a commodity now. 
What you do with them is not.

Ready for market research -> manufacturer handover node based workflow? 

We're cooking something massive and we are going to connect the best tools out there to make fashion AI/3D workflow easy, fast and scalable like never before!

PS. We're already connected with some great tools providers but if you are working in 3D/AI fashion space and want to get on board we are happy to discuss!

First step for garment feasibility? Existing fabric. 

You can't cut from something that doesn't exist. 
You adjust your pattern to fabric properties. 
Not the other way around.
 
When you start from an idea and not from fabric, 
you risk disappointing yourself.
 
→ You won't find the right one. 
→ The right one will cost too much. 
→ It's impossible to achieve the result you're looking for.
 
That's why we developed the "fabric matching" node at fashionINSTA.
 
You generate your idea. 
Fabric matching looks for fabric that ACTUALLY EXISTS.
 
It returns: 
→ Full description
→ Fabric properties
→ Link from the fabric shop
→ Fabric swatch for your reference
 
Then you regenerate your design with the real fabric.
 
Now you can buy it and cut the pattern. 
Now you can calculate the garment cost. 
Now you can match it against your margins.
 
Before you fall in love with an idea that can't exist.
 
Next steps we're building?
 
→ Full BOM 
→ Cost calculator 
→ Pattern export
 
What would you add? 
How would you improve this workflow?
 
Drop your ideas below!

PS. Subscribe to our waitlist to be among the first to try it!

How many AI subscriptions do you have?
I had 36 in 2025. Way too much. 
I cut them to 5 in 2026.

In the startup world, 
they are telling you to focus on ONE problem.
They tell you to build for ONE ideal client.
They tell you to ship fast and iterate.

And that's fine.
It's the same thing we did with fashionINSTA.

It's the only way to validate you have a market before you spend hundreds of thousands on useless development.

But, as a user, I pay for 30+ subscriptions because all those tools solve only ONE problem I have. 

They are often solving also others, I don't care about, but I have to pay for.

And it is an issue.
I have to move inputs/outputs across all of them.
I have different links, different passwords, different bills.

It's annoying.
It's expensive.
It's wasting my time.

I was always against superapps.
The ones that lock you in, 
the ones that are not interconnected.

For admin stuff,
For marketing purposes,
For sales and CRM, I built my own tool.

Exactly as I need and as I want it.

I pay just for the LLM call when I use it.
No subscription, no setting up several tools,
No moving input/output back and forth.

But what about fashion workflow?

There are amazing tools like Weavy or Flora, that are bringing huge value - you are able to use several tools in one place and pay for calls you make.

But they cover just one part of the process, the one related to visual market research. Eventually, the marketing one if you start from existing garments,

What about everything in the middle?

Tech specs
Patterns
PLM

Just to mention some.

What about 3D assets you can change only in ONE software, since the tools aren't connected?

What about all the time you lose moving between tools?

fashionINSTA is different here. 
We are not trying to build the wheel.
We don't have to replace tools that work.

We connect them in a single workflow that takes you from idea to market.

Think Weavy, but adding a production layer.

Think Lovable when you prompt your workflow and build from the bricks you already have.

Think real-time info about AI tools you ACTUALLY USE so you can turn off subscriptions you don't need.

Which tools would you like to connect in your fashionINSTA workflow?

In 6 years, the NFT hype is gone.
 
Will generative AI images have the same end?
 
I've just learned that NFT Paris won't happen this year.
Unfortunately, the wide use of NFTs didn't stick.
Such an important sign. And the end of an era.

Many bought the assets.
Many built companies around it.
Many invested in the technology.
Many bet their entire careers on them.
 
But there was too much speculation around technology that actually has great use cases.
 
Just not the ones that went viral.
 
Not digital ownership of goods that don't exist in real life.
Not tools too complicated to be adopted widely.
Not Bored Apes & Co.
 
Random folks don't want to be hackers. 
Customers don't learn entirely new tech for minimal gain. 
Educating customers is the hardest path you can choose.
 
If there are no immediate rewards, new tech cannot be adopted widely.
 
Is this teaching us something?
 
A lot.
 
Think about AI.
 
It's cool. It's fun. 
It's easy to start with. 
And it can be pretty useful.
 
It spread like a virus.
 
But now we're at the point where we need to find ways to earn from it.
 
And to earn, we need tangible results.
 
I don't speak only of money. 
Sometimes it's hard to quantify in dollars from the start.
 
I speak time. 
I speak output that brings results. 
I speak companies that are WILLING TO PAY for the output.

I speak people who can do their job 
→ faster
→ better 
→ cheaper
Or all of the above.
 
It works amazingly with coding.
 
I'm cancelling my tool subscriptions one by one as I build my own AI-based tools and agents.

→ PR
→ sales
→ marketing
→ all admin stuff
→ investor relations
→ knowledge database
→ social media analysis
→ custom website builder
→ node-based automatic workflows

They're cheaper. 
They do exactly what I want. 
How I want it.
 
For me, hyperpersonalized apps are the future of business. And it's here now.
 
Can we do the same with AI in fashion?
 
I don't think we're there yet. 
But I see many people looking in this direction. 
And I'm really curious how it will unfold.
 
The alternative?
 
Going back to business as usual. 
And finishing the genAI era in fashion the same way the NFT one finished.

We developed something a bit scary.
But this is where AI is going.
 
fashionINSTA has autonomous agents. 
We call them smart nodes.
 
The difference from standard AI workflows?
Standard nodes: input → instruction → output
Smart nodes: input → learning/validation loop → output
 
They get better with time.
They learn from their mistakes.
They find flaws in their own reasoning.
They learn from your input and feedback.
 
When I look at the logs, I'm honestly a bit scared.
 
AI talking to itself:
"I made a mistake here. I should do this instead."
"Last time I did it this way, the output was great. Let's try that again."
 
It brings much higher precision.
 
But it also brings extreme personalization.
 
With time, your tool learns your preferences.
Your style. Your taste. Your standards.
 
And delivers output that feels like yours.
Not generic. Not "close enough."
Yours.

Already cooked and now seasoning... learning!:
→ Fabric research node
→ .DXF pattern node
→ Nesting node
→ Pricing node
→ E-commerce photoshoot node
→ Colourway node
→ Tech pack node
 
Want to try it? 
Let me know in the comments.

taplio

The Battle

Shahab Shirmohammadi

Sylwia Szymczyk

Followers

Posts

Engagements

Likes

Performance ScoreSylwia Szymczyk's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Shahab Shirmohammadi

Sylwia Szymczyk

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Sylwia Szymczyk's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreSylwia Szymczyk's LinkedIn Performance