back to project
At the beginning of the branding process, all I had were the product photos we took and the logo.
There weren’t that many photos, and they didn’t cover every angle.
the first draft of the logo
the logo after the first round of critique
the final version of the logo
typography pattern
hand drawing pattern
The drawing shows the tools and process we used to make the bag it products.
I really like how M/M Paris combines hand-drawn elements with graphic design. It fits well with the brand positioning of something like bag it, which is a handmade luxury product—so I created the drawing on the left with that in mind.
This was the first version of bag it back in 2020. At the time, I could only design using the existing patterns, logo, and product photos.
01
02/28/2025
I got laid off.02
This also marked a new beginning—I started rebuilding my portfolio from scratch.03
I've always felt that bag it had a lot of potential, so I decided to remake it using the skills and tools I have now—kind of like how classic games get remastered with modern technology.04
So the AI tools and Photoshop skills I picked up at my last job ended up playing a big role in this project. The job itself wasn’t great, but without that experience, this bag it reboot probably wouldn’t have happened.
I decided to start by generating model images with AI—this was in April 2025, right after ChatGPT released its image generation feature in March. I thought it was the perfect time to try using it to create AI models that could showcase my product.
But then in April, ChatGPT placed restrictions on that feature—AI-generated people couldn’t interact with real-world objects. Basically, no matter how realistic the model looked, it couldn’t be shown holding or using something from the real world.
I couldn’t just send in a model image and my bag it product photo and ask it to generate a combined image of the model holding the bag.
So I had to get creative and figure out a workaround.This is a reference photo I found of a model. I sent it to ChatGPT and asked it to describe the image and turn that into a MidJourney prompt.
This is one of the images MidJourney generated based on the prompt from ChatGPT.
When I asked ChatGPT to write the prompt, I added a few specific elements: I wanted a forest background and a plastic bag floating in the air. That strong contrast between nature and the plastic bag felt visually striking to me—and it also ties back to bag it’s sustainability theme.
I also included a prompt for the model to wear a hat, which would save me some editing time later on when compositing.
Step 1
I started by using Photoshop’s Generative AI to remove the model from the image. At that time, ChatGPT couldn’t even handle removing real people from a photo while keeping the background intact—not sure if that’ll change in future updates.
Step 2
Since ChatGPT wouldn’t let real people wear real hats, I figured—why not just have a line-drawn figure wear one instead?
So I quickly traced the model in the image on my iPad.Step 3
I gave ChatGPT the prompt: “Put the bucket hat from image 123 onto the line-drawn figure.”
And it worked!Step 4
I used Photoshop to place the hat generated by ChatGPT onto the model’s face.
The great thing about this workflow is that the hat already comes with built-in lighting and shadows, which makes it feel more natural.
Also, since I only had a limited number of product photos and couldn’t cover every perspective, I needed AI to generate product images that matched the angle and perspective of the model shots.
MidJourney happened to generate this image where the model’s face is covered by a floating plastic bag—and I found it incredibly powerful. When a face is obscured like that, it forces people to notice the plastic bag as something unnatural—something that doesn’t belong, that can’t simply break down or disappear.
Strictly speaking, this is my first true personal graphic design project since 2020. It’s been five full years since I seriously created something just for myself.
The way I used to work before the pandemic no longer applies. Back then, I would do tons of research until I had a clear design visual in my head—then the execution part was just about making it real.
But that approach doesn’t quite work here. First, I haven’t come across many fashion branding projects lately that feel truly fresh or exciting. Second, AI is unpredictable—I never know what it’s going to generate, whether it’ll be surprisingly brilliant or totally off.
This image is a perfect example. It was a random output from MidJourney, but it sparked a whole new creative direction. I ended up using the concept—plastic bags obscuring the model’s face—in all the following images, and the results turned out great.
Now, my process is less about having a fully-formed, super exciting idea in my head and more about starting with a vague notion—then letting the process, AI, and unexpected visual discoveries guide me toward the final result.
In a way, that actually fits how I’ve always worked. Every one of my projects has been a step out of my comfort zone. After all these years, my work has become so unique that it’s hard to innovate just by combining references or existing styles. With bag it, there really wasn’t a clear visual precedent—I hadn’t seen anything quite like it on the market. So there was no way I could’ve envisioned it from the start.
Which means, for this one, I had to feel my way through it.At first, I was just randomly testing out model images I found on lundlund.com, feeding them into AI to see what would come out. But the vibe never really landed—it wasn’t giving that Asian-underground-cool energy I was going for.
Then I came across a photo by Thomas Cooksey featuring an Asian male model. That one hit different—the vibe finally clicked.
But I didn’t realize right away that I had found the right direction. It wasn’t until I laid out all the generated images side by side and felt something was off—none of them felt cohesive or quite “right.”
That’s when I saw this particular set of images again, and it hit me: this was the look I had been searching for all along.
Ask ChatGPTAfter that, I started exclusively using Thomas Cooksey’s photography as the base for AI generation—I’m genuinely a big fan of his style.
What I realized was that instead of blindly trying random images and hoping for the best, I needed to first identify a photographer whose aesthetic matched bag it's brand. Then I could use prompts inspired by their work to generate model images that actually made sense for the project.
So thank you, Thomas.
From the beginning, I really wanted to feature Asian or Japanese models. Japan, as a country that deeply values sustainability, aligns perfectly with bag it's ethos. But early on, I didn’t know how to consistently generate AI images with both the right aesthetic and good-looking models—so I temporarily let go of that idea.
I’m really glad I found my way back to it.
After generating all the model images, I moved on to exploring the visual direction for bag it. But something still felt off. The models looked right, the vibe was there—but the graphic design side wasn’t landing.
Here are a few of the early experiments I tried.
The turning point came when I stumbled upon a really interesting brand on Behance—it was a project called WCBMO by designer Айдентика.
Inspired by that, I redesigned my bag it logo and started layering it with phrases in different typefaces and even chemical formulas.
And just like that—it finally felt right.
Before diving in, I made a list of all the design assets I’d need for the brand—things like business cards, posters, print materials, social media pages, invitations, price tags, receipts, and so on.
Once I had the right model images and the logo direction nailed down, I decided to start with the business card. It’s a relatively simple piece, and the layout from Айдентика’s WCBMO project gave me a great starting point.
01
After finishing the business card, the visual identity of bag it finally clicked into place. From that point on, everything just flowed naturally.
Below is a screenshot of all the design assets I created.
02
The actual design work only took about four days—but the early-stage exploration, figuring out how to work with AI and locking down the visual style, took around a month and a half.
These designs didn’t make the final cut because they didn’t quite fit with the overall bag it aesthetic.
I believe that as a creator, sharing your creative process is really important—and it can be inspiring for other creators too. I personally love seeing behind-the-scenes stories of how different works are made, and they’ve influenced me a lot.
So I spent another three hours putting together this process page. Hope you enjoy this project!
Revisiting and remaking bag it has meant a lot to me on a personal level.
After 2020, I started to lose my creative spark—and for a designer, that’s a devastating thing.
I pushed through two tough years of grad school at SVA and graduated with a lingering sense of regret. Without that creative drive, I hadn’t made anything I truly felt proud of—whether in motion or graphic design.
After graduating, I joined Acquco, where I mostly worked on Amazon product images. It was far from the kind of design work I had always dreamed of doing.
So when I got laid off, I started rebuilding my portfolio—and that’s when bag it was born.
Strangely enough, once I chose to believe that I was still a good designer, the inspiration started to come back.
My process had changed—now, it didn’t start with a fully-formed, exciting idea in my head. Instead, I began with a vague concept and let AI tools or beautiful new visual references guide me. The creativity unfolded through exploration. And that was enough to bring me closer to the final solution.
And I’m really proud that I made it happen.
I still remember back at MICA, right before graduation, the school brought in industry designers to critique our portfolios. One of them said something that stuck with me:
“This might be the last time you ever design something just for yourself. Once you start working, that window closes.”
At the time, I didn’t quite believe them—and now, I’m glad I didn’t.
Because bag it proves they were wrong.
Design is more than a job to me. It’s something I truly love.
Creating is something I’ll keep doing for the rest of my life—not for money, but because it feeds my soul.
That’s all there is to it.
back to project