Sensei - Building an AI-native UX engine

Sensei - Building an AI-native UX engine

Role
Co-Founder · Design Engineer
Responsibility
Strategy · Product · Design Engineering
Timeline
2025 — Present
Outcome
Live product
Role
Co-Founder · Design Engineer
Responsibility
Strategy · Product · Design Engineering
Timeline
2025 — Present
Outcome
Live product

I - The problem

I - The problem

AI has made it remarkably easy to build software, and even easier to make bad software.

I built Sensei to close that gap. It's an AI-native UX auditing platform that evaluates digital experiences against established design principles and a library of 200+ industry benchmarks, turning what would traditionally be hours of manual review into a structured assessment in minutes.

The ambition was to make expert-level design critique accessible without reducing it to another generic AI-generated opinion.

AI has made it remarkably easy to build software, and even easier to make bad software.

I built Sensei to close that gap. It’s an AI-native UX auditing platform that evaluates digital experiences against established design principles and a library of 200+ industry benchmarks, turning what would traditionally be hours of manual review into a structured assessment in minutes.

The ambition was to make expert-level design critique accessible without reducing it to another generic AI-generated opinion.

II - Designing the experience

II - Designing the experience

The central UX challenge was translating complex analysis into something immediately understandable and actionable. Early iterations were comprehensive but overwhelming, burying useful insights beneath layers of information.

I restructured the experience around a clear hierarchy: an immediate assessment, prioritised issues, and deeper recommendations for those who need them. Rather than presenting a wall of AI-generated text, Sensei uses a visual language of scores, evidence and contextual feedback to help users understand where to focus.

III - Design intelligence

III - Design intelligence

What began as a single-page auditing tool evolved into a broader product capable of evaluating entire websites. Site Audits introduced new UX challenges around page discovery, progress, navigation and presenting findings across multiple interconnected experiences.

I designed the end-to-end flow and built the product using AI-assisted development, moving directly between interface design and implementation. Working this way allowed me to test decisions in production rather than relying exclusively on prototypes, while keeping the underlying complexity largely invisible to the user.

IV - Shipping Sensei

IV - Shipping Sensei

Sensei V1 launched in September 2026, bringing page audits, multi-page analysis, persistent reports and a dashboard into a single product.

Beyond the product itself, Sensei became an exercise in designing and engineering an AI-native experience from the ground up, where the real value isn't simply what the model can generate, but how thoughtfully that intelligence is translated into a useful piece of software.

Sensei V1 launched in September 2026, bringing page audits, multi-page analysis, persistent reports and a dashboard into a single product.

Beyond the product itself, Sensei became an exercise in designing and engineering an AI-native experience from the ground up, where the real value isn’t simply what the model can generate, but how thoughtfully that intelligence is translated into a useful piece of software.

samuel.e.mclean@gmail.com

samuel.e.mclean@gmail.com

samuel.e.mclean@gmail.com