Artisk

Gen AI brand visual asset management platform

  • Project type: B2C Web App

  • Role: UX Research, UI/UX Designer

  • Industry: Design. AI, Technology

  • Duration: Summer 2024 - Winter 2024

  • Project type: B2C Web App

  • Role: UX Research, UI/UX Designer

  • Industry: Design. AI, Technology

  • Duration: Summer 2024 - Winter 2024

Artisk is an AI design platform that makes creating visuals really simple, even for people without a design background. It lets users quickly turn ideas into professional logos or posters, and build a design system based on that, so anyone can create with confidence.

Click here to try Artisk

Artisk is an AI design platform that makes creating visuals really simple, even for people without a design background. It lets users quickly turn ideas into professional logos or posters, and build a design system based on that, so anyone can create with confidence.

Click here to try Artisk

Design Outcome

❓Problem

  • Design for users with no design background

  • Existing design tools are too complex and time-consuming

  • Limited time & budget

❓Problem

  • Design for users with no design background

  • Existing design tools are too complex and time-consuming

  • Limited time & budget

👑 Goal

  • Increase platform retention and attractiveness

  • Integrate with the existing logo generator

  • Ensure the UI/UX is clean, consistent, and low-barrier for non-designers

👑 Goal

  • Increase platform retention and attractiveness

  • Integrate with the existing logo generator

  • Ensure the UI/UX is clean, consistent, and low-barrier for non-designers

Design Process

Market Research

For Artisk, one of the main ways we understood user needs was through competitive analysis. Since we were a startup, we didn’t have the resources to conduct many user interviews, so we used competitor research to fill that gap — learning what worked well and where users still struggled.

When we did market research for Artisk, we grouped tools into three categories. First, professional tools like Adobe Photoshop and Illustrator. These are still the industry benchmark, and Adobe is also pushing AI features like Generative Fill. But the problem is accessibility — for non-designers, the learning curve is way too high, the cost is significant, and even installation can be a barrier.

Then we looked at Canva. This is closer to where Artisk wants to be: easy, intuitive, template-rich, and perfect for small business owners with no design background. But Canva also has its weaknesses: because it’s made for everyone, many professional features are simplified or missing, branding consistency can be hard to maintain, and their AI features are buried deep in the design process, making them difficult to discover.

Finally, we analyzed new AI-driven tools like Kittl, Flair.ai, Sivi, and Pencil. These are fast and great at generating branding assets or ad creatives, but their focus is much more on advertising performance than on design quality. Their scope is also narrow and the outputs often lack consistency.

What we learned from this research is that there’s still a gap for a tool that combines the usability of Canva, the power of AI, and the design consistency of a system like Adobe — but in a way that’s simple, affordable, and accessible to non-designers. That’s where Artisk positions itself.

Market Research

For Artisk, one of the main ways we understood user needs was through competitive analysis. Since we were a startup, we didn’t have the resources to conduct many user interviews, so we used competitor research to fill that gap — learning what worked well and where users still struggled.

When we did market research for Artisk, we grouped tools into three categories. First, professional tools like Adobe Photoshop and Illustrator. These are still the industry benchmark, and Adobe is also pushing AI features like Generative Fill. But the problem is accessibility — for non-designers, the learning curve is way too high, the cost is significant, and even installation can be a barrier.

Then we looked at Canva. This is closer to where Artisk wants to be: easy, intuitive, template-rich, and perfect for small business owners with no design background. But Canva also has its weaknesses: because it’s made for everyone, many professional features are simplified or missing, branding consistency can be hard to maintain, and their AI features are buried deep in the design process, making them difficult to discover.

Finally, we analyzed new AI-driven tools like Kittl, Flair.ai, Sivi, and Pencil. These are fast and great at generating branding assets or ad creatives, but their focus is much more on advertising performance than on design quality. Their scope is also narrow and the outputs often lack consistency.

What we learned from this research is that there’s still a gap for a tool that combines the usability of Canva, the power of AI, and the design consistency of a system like Adobe — but in a way that’s simple, affordable, and accessible to non-designers. That’s where Artisk positions itself.

Internal Discussion - User Flow & Potential Features

After market research, we have an interal discussion for our user flow and potential features. We first mapped out the user flow. A user either comes in with a design system and uploads a logo, or they don’t, in which case AI helps them generate one. That logo then feeds into future design projects.

When they move to the poster generator, the flow is simple: they pick the purpose and ratio, answer a few quick questions, and we generate four poster options. They choose one, edit details, and save.

Prototype - Low fi

Iteration

  1. Redesign Poster Ratio Selection

In the original user flow for selecting poster ratios, the design presented a clear and straightforward list for users to choose from. However, during team discussions, we identified the need for a more flexible and iterative approach that better integrates with the overall design system. The new layout replaces the list with a horizontally grouped arrangement of clickable buttons, maintaining clarity while improving the user’s browsing and selection experience by reducing cognitive load.

  1. Redesign Poster Ratio Selection

In the original user flow for selecting poster ratios, the design presented a clear and straightforward list for users to choose from. However, during team discussions, we identified the need for a more flexible and iterative approach that better integrates with the overall design system. The new layout replaces the list with a horizontally grouped arrangement of clickable buttons, maintaining clarity while improving the user’s browsing and selection experience by reducing cognitive load.

  1. Redesign Poster Ratio Selection

In the original user flow for selecting poster ratios, the design presented a clear and straightforward list for users to choose from. However, during team discussions, we identified the need for a more flexible and iterative approach that better integrates with the overall design system. The new layout replaces the list with a horizontally grouped arrangement of clickable buttons, maintaining clarity while improving the user’s browsing and selection experience by reducing cognitive load.

  1. Layer panel UI redesign

Initially, the panel was large and non-intuitive, as it didn't resemble an editable interface. Following feedback from multiple users during testing, I redesigned the panel to be thinner and more streamlined and user are able to turn on this panel to edit the details.

  1. Layer panel UI redesign

Initially, the panel was large and non-intuitive, as it didn't resemble an editable interface. Following feedback from multiple users during testing, I redesigned the panel to be thinner and more streamlined and user are able to turn on this panel to edit the details.

  1. Layer panel UI redesign

Initially, the panel was large and non-intuitive, as it didn't resemble an editable interface. Following feedback from multiple users during testing, I redesigned the panel to be thinner and more streamlined and user are able to turn on this panel to edit the details.

3. Enhancing the User Experience for Adding New Layers and Poster Elements
After improving the visual clarity of the layer panel, I started to think more deeply about how users interact when adding new layers. My initial idea was to place the “Add Layer” function in the top navigation bar, but that solution turned out to be limited: it wasn’t discoverable enough for beginners and didn’t scale well for future features.

Through iteration, I explored a more holistic approach. First, I considered different user scenarios: beginners need a prominent “+ Add Layer” entry point, while advanced users benefit from shortcuts and contextual menus. Second, I thought about edge cases, such as how to prevent clutter when many layers are added, and providing undo/confirmation flows for accidental clicks. Third, I designed with scalability in mind: instead of a single “add” action, the new toolbar layout can expand to include text, image, and AI-generated components in the future.

3. Enhancing the User Experience for Adding New Layers and Poster Elements
After improving the visual clarity of the layer panel, I started to think more deeply about how users interact when adding new layers. My initial idea was to place the “Add Layer” function in the top navigation bar, but that solution turned out to be limited: it wasn’t discoverable enough for beginners and didn’t scale well for future features.

Through iteration, I explored a more holistic approach. First, I considered different user scenarios: beginners need a prominent “+ Add Layer” entry point, while advanced users benefit from shortcuts and contextual menus. Second, I thought about edge cases, such as how to prevent clutter when many layers are added, and providing undo/confirmation flows for accidental clicks. Third, I designed with scalability in mind: instead of a single “add” action, the new toolbar layout can expand to include text, image, and AI-generated components in the future.

3. Enhancing the User Experience for Adding New Layers and Poster Elements
After improving the visual clarity of the layer panel, I started to think more deeply about how users interact when adding new layers. My initial idea was to place the “Add Layer” function in the top navigation bar, but that solution turned out to be limited: it wasn’t discoverable enough for beginners and didn’t scale well for future features.

Through iteration, I explored a more holistic approach. First, I considered different user scenarios: beginners need a prominent “+ Add Layer” entry point, while advanced users benefit from shortcuts and contextual menus. Second, I thought about edge cases, such as how to prevent clutter when many layers are added, and providing undo/confirmation flows for accidental clicks. Third, I designed with scalability in mind: instead of a single “add” action, the new toolbar layout can expand to include text, image, and AI-generated components in the future.

Solution

Final Design

Impact & Reflection

The impact was clear: both users and our internal team found the new poster generator much simpler and more intuitive. It also increased the 5% of the retention and platform’s attractiveness, because users could now generate both logos and posters in one place and this place will also mermorize their design system, without switching tools.


For me personally, I learned a lot about designing under business constraints — for example, balancing user needs with time, budget, and technical feasibility. I also improved my collaboration with developers, and got better at prioritizing features: knowing which UI elements to optimize now, and which ones to defer.


At the same time, there were limitations. For example, we didn’t run large-scale usability testing yet. And customization is still somewhat limited I believe those are things that could be improved in later iterations. Also, in a real product release, performance and loading speed would be critical areas to improve.


Compared to my academic project, which was more idealized, Artisk felt much closer to a real business product. It gave me experience bridging theory and practice, and taught me how to balance usability, business goals, and technical feasibility in a real-world setting.


If you'd like to learn more about Artisk, please click the link below to access Artisk's official website.

Learn More

← Back to Projects
SheWell →