top of page

LENSA AI

Frame 32.png
Frame 873.png

UI/UX, PRODUCT RESEARCH, USER TESTING

A Seamless AI-Powered Visual Experience with Lensa

Frame 31.png

My Role

UI/UX Designer, Product Owner

Team

6 members

Duration

2 weeks

INTRODUCTION

When AI Meets Art

In 2023, Lensa’s Magic Avatars were everywhere.
 Millions of users uploaded their selfies to get stylized portraits powered by generative AI. It was fun, fast, and wildly viral.


But as the hype grew, so did the questions:
What happens to my data?
Can I trust what the AI creates?
Why doesn’t this look like me?

 

In a space where technology moves fast, trust often lags behind. That’s where our team came in.

2f5204b1-5471-4ce3-9c95-375c2fc5d239.gif
add-profile-picture.png

THE OBJECTIVE

Empowering Users to Create Stunning Visuals with AI

The objective of this project was part of my Master’s capstone collaboration with IBM, where we focused on designing more ethical, user-centered AI experiences. Our mission was to redesign the Magic Avatar experience to foster greater transparency and trust, while maintaining the magic and creativity that users love. By prioritizing user needs and ethical considerations, we aimed to build an AI tool that feels intuitive, trustworthy, and fun, without compromising its creative potential.

BUSINESS IMPACT

Why Visibility was a SaaS Imperative

To ensure the redesign was impactful, we aligned user-centered goals with Lensa AI’s broader business objectives. By creating a more ethical and user-centered experience, we positioned Lensa AI to not only meet evolving user expectations but also stand out as a leader in responsible AI innovation.

coffee-tea.png

Strengthen Brand Loyalty

Building transparency and trust fosters deeper emotional connections with users

about-our-team.png

Increase User Retention

A more ethical, user-centered experience keeps users coming back

chart.png

Drive Long-Term Growth

Aligning with responsible AI practices positions Lensa AI for sustainable success and market leadership

THE PROBLEM

Despite the appeal of AI-generated visuals, users struggled with a lack of transparency and control, leading to confusion, diminished trust, and growing concerns around data use — ultimately putting the brand’s credibility at risk.

THE SOLUTION

We took an ethical design approach to reimagining the Magic Avatar experience — empowering users with transparency, informed choices, and a clearer understanding of AI, without compromising the creativity and delight that made the feature special.

DISCOVERY

What We Discovered Beneath the Magic

Through secondary research, we identified key challenges users faced, including a lack of human control, transparency, explainability, and concerns around bias and privacy. These insights helped us pinpoint critical pain points in the user experience. From there, we generated "How Might We" statements to guide our brainstorming and develop meaningful solutions.

660b2315327bc22fc176703b_hmwlensa.png

USER RESEARCH

Who's using Lensa AI?

To stay user-centered, we created a persona that captured the motivations and concerns of our target audience. Bev Mov helped guide our design choices, ensuring we addressed real user needs like transparency, privacy, and creative control.

6608d5a78f15f22593bc1a82_UX Persona.png

COMPETITIVE ANALYSIS

What We Discovered Beneath the Magic

In our research, we explored how competitors address user concerns around privacy, transparency, and feedback. We found that while many apps provide basic solutions in these areas, there are opportunities for improvement in user control and clarity.

Key Observations
🔒
👁️‍🗨️
📝

Privacy

Competitors explain how data is collected and provide options for users to opt out. Bullet points are commonly used to clarify the purpose of data collection.

Transparency

Competitors make it clear why the system made certain decisions and are upfront about potential risks. Many apps use Explainable AI (XAI) to offer transparency about their AI systems' actions.

Collecting Feedback

Common feedback options include thumbs up/down, content reporting, and filtering. Some apps also offer like/dislike buttons to capture user preferences.

create-account.png

SOLUTIONING

Key Features & Solutions

Pain Point

Frame 872.png

Solution

USER FLOW

Implementing Solutions into the Current App

Using a user flow, we identified key problem areas where our solutions (represented by the red postcards) could have the most impact.

660c9f16f332344c572eaed5_User Flowuser flow.png

ROADBLOCKS

Designing for Consistency and Clarity

For the high-fidelity mockups, I collaborated with the UI team to refine wireframes and ensure they adhered to Lensa's existing branding guidelines. This was crucial in maintaining design consistency while creating a user-friendly experience. 

660b2a4526f54b4b3075b2ac_solution.png

TESTING & REFINEMENTS

Discovering user sentiment on our Lensa AI re-design

To measure user sentiment toward our redesigned Lensa AI experience, we conducted two rounds of moderated usability and desirability testing with 16 participants. Using a controlled vocabulary list and benchmarking against KPMG’s "Trust in AI" study, we aimed to boost user trust and identify usability issues. 

6% increase of users willing to trust Lensa's AI

Based on a 95% confidence level as shown in our graph, users are more willing to trust our version of Lensa AI technology in comparison to our benchmark, describing the experience as more customizable and straightforward, though some still found it suspicious.

6608e23e2a64b123bedb9262_image (2).png
660ca13007f4b61202c82f87_Screenshot 2024-04-02 at 5.20.16 PM-p-500.png
Evolving the Design Through User Feedback

Based on insights from usability testing, we refined our designs to better build trust and clarity. We improved explanations of how the AI works, adjusted tone and added helpful links for transparency, and applied the principle of feedback to keep users informed throughout their journey.

Explaining how the AI works

660af21d2df5a5cb2151447c_1.png

Changing tone and including links

660b0136036fb64d6f6c0449_2.png
660b0135900ec44917dc3f6b_3.png

Applying principle of feedback

660b0889c0cd7e1b415ebc8a_Frame 1050.png
660b0888211bb85e70898416_Frame 1051.png
660b0889ede4a339ee5046af_Frame 1052.png

DELIVERY

Final Designs

Frame 322.png
Explain the AI Design Process

An informative video shows how and why the Magic Avatar generates its results

Group 38.png
Transparency

Full disclosure on how personal information is being used and giving users power to opt in or opt out

AI Awareness

An informative video shows how and why the Magic Avatar generates its results

Feedback

Gathering user feedback is the best method to understand what the users want and to improve the same in the app

golden-shimmery-stars-pattern-black-background.png

CONCLUSION

Reflecting Forward

Advancements in AI technology bring both incredible potential and new challenges. As AI continues to expand, ensuring systems are trustworthy, ethical, and transparent is more important than ever. The Lensa redesign allowed us to tackle trust issues head-on and contribute to the ongoing conversation about the responsible use of AI.

⚡ Challenges

Working within tight time constraints forced us to prioritize certain tasks, balancing user needs with business goals. Scope limitations and feasibility considerations shaped our decision-making process.

🚀  What's Next

Though we couldn’t implement every solution, we recognize the value in exploring additional opportunities to further enhance transparency and user control in AI-powered tools.

ChatGPT_Image_Apr_24,_2025,_01_10_03_PM.jpg

ERICA DOO

© 2025 Erica Doo - Thoughtfully designed, gently told.

bottom of page