AI can help us reduce cognitive load and surface information that helps our user achieve their goal, complete their need and go back to their lives. My focus is on making new technology updates feel intuitive, trustworthy, and human-centered.

I approach AI design with three guiding principles:

  • Contextual value : AI should enhance the experience by surfacing the right information at the right moment, never adding noise or complexity.

  • Trust through transparency : Users need to understand why an AI recommendation was made and have the ability to adjust, override, or dismiss it.

  • Adaptive learning : AI should evolve with user behavior, but always within safe, predictable boundaries that respect privacy and compliance.

My process combines human-centered research, rapid prototyping, and iterative testing, ala old school UX to uncover how people interact with adaptive systems. By pairing design intuition with data and technical constraints, I create AI-driven flows that feel helpful, reliable, and empowering, never introducing experiments at the user’s expense.

How my design process adapts for AI

My process is iterative, and non-linear, but well established. 💎 The steps in my design process have updated to include AI needs.

Start here!

💎 Frame the problem & opportunity

 I begin by clarifying the user need and identifying where AI can add real value: whether that’s reducing cognitive load, personalizing an experience, or anticipating next steps. I map this against product goals and constraints to ensure AI is solving the right problem.

💎 Landscape research and understand signals:

I explore how people currently approach the task and what level of automation or assistance they are comfortable with. In AI design, understanding comfort zones, transparency needs, and boundaries of trust is as critical as usability.

💎 Define human and AI roles

I map workflows to decide what the AI should handle versus what the user should control. This includes defining fallbacks for when the AI is uncertain, ensuring the user never feels trapped or misled.

💎 Prototype adaptive flows

I design and prototype the flows that showcase AI responses change based on context. I pay special attention to explainability cues (e.g., “Because you did X, we recommend Y”) and controls for user override.

💎 Test!

I validate early and often running usability tests to measure trust, comprehension, and perceived value. For AI features, I evaluate not just whether users can complete a task, but whether they feel confident and in control while doing it.

💎 Iterate

I refine the design using both qualitative feedback and quantitative signals (clickthrough, dismissal rates, engagement). This dual lens helps balance user trust with system accuracy.

💎 Collaborate with engineering, data et al.

I work closely with engineers and data scientists to understand model limitations, bias risks, and latency. This ensures the design reflects the real capabilities of the AI and avoids overpromising.

💎 Evolve & scale with systems thinking

All successful patterns are, ideally, integrated into the design system, creating reusable AI-friendly components (such as recommendation cards, adaptive tooltips, or confidence indicators) that can scale across products.