.png)
In the first installment of our new “AI by Design” series exploring the innovative ways in which design leaders are integrating AI into their work, we spoke with Nanda Gopal, Principal Designer at Zemoso Technologies, an agency working with start-ups and entrepreneurs to bring their product visions to life. Nanda is also a seasoned mentor at Designlab, working with students across several of our courses.
We spoke with Nanda as part of our outreach for our recent AI in UX & Product Design Survey, where we surveyed over 120 working design professionals about how they’re adopting AI in their UX workflows. Nanda has developed a comprehensive three-phase approach to AI adoption that has significantly accelerated their team's capabilities.
What makes Nanda's approach particularly interesting is how systematically he’s implemented AI across their entire design process—creating distinct workflows for research, execution, and handoff phases. This structured adoption not only boosts efficiency but also offers valuable insights into the evolving role of designers in an AI-augmented world.
Breaking Down AI Integration into Three Phases
Nanda's team began experimenting with AI tools immediately following ChatGPT's initial launch, developing a structured approach that spans the entire design workflow:
Phase 1: Research
In the research phase, Nanda's team uses tools like Perplexity, Notebook LM, and ChatGPT—each selected based on specific use cases.
"For example, if you have [existing design research, information, or documentation] to work off of, it's preferable to go into Notebook LM because that way we can ask more targeted questions and limit it to our sources. But if it's a more open-ended starting point, then web-based research with tools like Perplexity would be the choice," Nanda explains.
This approach has been particularly valuable for an agency like Zemoso which frequently works across diverse domains. By developing AI-based information libraries, they've created a system where even junior designers can quickly gain sufficient knowledge before engaging with stakeholders.
This efficiency has had a tangible impact, with Nanda noting "What used to take us about two to three days, we were able to bring it down to like a day, day and a half. Which is a significant gain."
This acceleration in the initial research phase allows the team to dedicate more time to strategic thinking and deeper problem-solving later in the process.
Phase 2: Execution
During the execution phase, Nanda's team uses tools like Claude, Creatie.ai, and Galileo to accelerate the design process—particularly during design sprints.
"The biggest hurdle we've had is that when you're working with entrepreneurs, especially in design sprint settings, you want to get as many ideas out as you can in front of the stakeholders, but designing each of those ideas is a time-intensive process."
Their solution? Designers sketch wireframes on paper, then use AI tools to generate HTML code from these sketches, which they import into Figma for refinement.
"Instead of starting at zero, I'm essentially using these tools to start at 50% or 60%, which means I just have to do that remaining polishing of the visual design. So I can get more variations out more quickly”, Nanda shared about this process.
This efficiency gain allows Nanda’s team to show clients additional options and variations. Instead of being able to show a client three of four variations on a design direction, Nanda’s team can now create double that amount without increasing the designer headcount.
While AI is allowing designers to rapidly iterate and present more options, Nanda feels that these tools are augmenting, rather than replacing, the creative output of the design team. "If you rely on AI, you get a very commoditized UI. So it's going to look similar everywhere, which means then you have even more chance to stand out if you can do something slightly different with the output”, said Nanda.
He believes that while AI excels at generating initial drafts, you still need a human designer to push design work to the point where it really stands out.
Additionally, there are still challenges with some tool limitations at this phase. Despite the impressive workflow they've established, Nanda still sees room for improvement in AI tools. Their wishlist includes better integration within design software like Figma:
"Once I make a component using AI and bring it into Figma, then doing the variations in the tool itself, instead of doing this multi-tool juggling, would be a really game changer for the team. Because I just need to live inside one tool."
Additionally, many of their workflows require designers to create wireframe sketches first, as Nanda has seen AI struggle with creating effective layouts. Not all designers are comfortable with sketching, so tools that could work directly from text prompts would further streamline their process.
Addressing these limitations would further empower design teams to leverage AI more seamlessly and effectively within their existing workflows.
Phase 3: Handoff
For the handoff phase, Nanda's team has developed custom GPTs to automate documentation—an often tedious but crucial aspect of the design process.
"One of the things designers hate doing is writing the documentation when you're handing it off to developers so that they have all the details necessary and they understand all the edge cases."
Their AI-powered solution generates comprehensive documentation that covers everything from padding restrictions to character limits, to how to deal with edge cases. These changes can cut the time designers spend writing documentation nearly in half, allowing designers to focus on more strategic and creative tasks.
Impact on Team Structure
When asked if AI has affected their staffing needs, Nanda noted a shift in their approach to design sprints:
"Typically, our model used to be three designers going to a design sprint. One of them is a senior, one is a mid-level designer, and then a junior. What we've noticed is you typically don't need the senior to be actively involved anymore if you're going through the workflow. The Senior designers just come in to review the output and ensure everything is to the desired standard."
However, Nanda emphasizes they're not yet at the point of dramatically reducing their design teams: "We are not in a place where we might fully remove all three designers and just replace it like one or two, just because these tools sometimes can fumble”.
With AI handling more of the execution and development aspects, Nanda believes the real value for designers will be at the intersection of product and design. "Your day-to-day document creations or your day-to-day initial first drafts of design will be accelerated. You can have a lot more bandwidth to have conversations with stakeholders about how a product fits into their larger context of a business, and solve for users’ perspectives rather than just pushing pixels.”
This shift suggests a future where designers spend less time on repetitive tasks and more on strategic thinking, user understanding, and aligning design with business goals.
Interested in learning more about how AI is transforming design workflows? Check out our AI for UX Design course to develop practical skills for integrating AI tools into your design process.
