Designing AI systems humans can trust.
I design systems where decisions matter. Not interfaces. Not features. Decisions.
Most products do not fail because the technology is wrong. They fail because the system makes it hard to know what to do next. Information is fragmented. Feedback is delayed. People are forced to hold too much in their heads at once.
I work in those environments. Clinical platforms. Safety systems. AI tools operating at scale. The goal is always the same: give people what they need to act with confidence, at the moment they need it.
Based in Dubai. Working across MENA and APAC for over a decade.
How I design
Turn complexity into signals
Systems that extract what matters from fragmented, high-volume information environments.
Structure how information flows
The sequence and timing of information often matters more than the information itself.
Support human judgement under pressure
Good design does not replace decisions. It makes them clearer, faster, and harder to get wrong.
The pattern
Across different domains, the pattern is often the same. Information is fragmented. Workflows are inconsistent. Decisions depend too much on individual interpretation.
The domain changes. Healthcare, safety, supply chain, AI. The structural problem rarely does. That cross-domain perspective is most of what I bring.
What I do for clients
UX strategy and vision
For product teams with a defined problem but no design direction.
I define what the product should be, not just how it should look. This means mapping the ecosystem, reframing the problem where needed, and producing a vision that teams can align around and execute against.
Decision system design
For founders and teams building AI-native or high-stakes products.
I design the logic, trust, and decision-making layer that determines whether people rely on your system or quietly stop using it. The interface is the last thing we talk about. The behaviour model, the failure states, and the trust architecture come first.
Research and validation
For teams who need to test direction before committing to build.
Field studies, journey mapping, and synthesis that moves organisations. I go to where the real problem is, from operational environments to clinical workflow analysis, and translate what I find into decisions.
AI and trust
I do not treat AI as a feature. I treat it as part of the decision system.
That means designing for calibrated trust. Not blind confidence in the model. Not scepticism that prevents use. A system where people know when to rely on it and when to question it.
The hardest part of AI product design is not the model. It is designing the moment between the output and the decision. That space is where most AI products succeed or fail.
If you are building something complex, where decisions actually matter, that is where I do my best work.
How we work together
Three ways to engage.
Short sprint
1 to 2 weeks. A focused problem, a clear output. Good for teams that need momentum.
Ongoing
Monthly retainer. Embedded strategic support across a longer initiative.
Embedded
Part of your team for a defined period. Best for complex problems that need sustained focus.
Thinking
Most AI UX problems are not AI problems
Many AI products fail long before the model does. The real issue is often unclear decisions, poor system feedback, and weak trust design.
When complexity is the product
In high-stakes systems, complexity cannot always be removed. The work is to structure it so people can act with clarity and confidence.
Trust is not the same as transparency
Showing more information does not automatically build trust. What matters is whether people can understand when to rely on the system and when to question it.