My Design Process and AI

AI is changing how we design and how our workflows evolve. Every designer develops their own way of integrating AI into daily work. After a period of experimentation and learning how different tools flow together, I have arrived at an approach that feels both practical and effective for my design practice.

Discover

In the discovery phase, I strongly believe in human creativity. AI plays a supporting role here by transcribing meetings and interviews. When preparing workshops, AI helps create agendas, plans, and templates. During workshops, I use AI to improve facilitation efficiency by summarizing discussions and synthesizing outcomes. Market research is also key at this stage, with exploration of opportunity spaces. Perplexity is currently my go-to tool for this.

Define

The definition phase is all about deeply understanding the target audience. I am cautious with synthetic users and prefer to combine in-person research with synthetic one. That said, AI-supported research is still better than no research at all. I also frequently use AI to analyze competitors, uncover patterns, and identify opportunities. (And competitor reviews and comments analysis is really a powerful and magic method to boost your usability). One method I have been using more lately is movie-industry-style storyboarding to shape compelling concepts and narratives.

Develop

In the develop phase, AI enables rapid exploration by generating multiple variants and directions. With Cursor and Figma Make, I can quickly build and validate prototypes. I also find it valuable to get an AI-driven critique of ideas before sharing them with colleagues. One of my favorite techniques is simulating attention and focus through generated heatmaps, ensuring that the most important elements receive the right level of attention, which still feels a bit magical. AI also shines here by taking over mundane tasks such as layer naming and visualizing component variants.

Deliver

In the delivery phase, I use AI to quickly generate documentation and create further content that aligns with the brand voice. Web analytics makes it easier to identify improvement opportunities across the journey and close the feedback loop. With sufficient traffic, it becomes possible to run highly nuanced experiments, create personalized experiences, and even predict behavior. TheyDo is currently the tool I use most for data-driven decision-making.

As tools continue to evolve, the real challenge is not adopting AI, but defining workflows and approaches that truly work for you. Creating synergy across multiple tools at different stages is where the real value emerges. Happy to share my experience in every step of the process and exchange perspectives on how others are integrating AI into their design workflows.