Zira
Production Dashboard
Role
Design Lead & ML Contributor
Year
2025
Disciplines
Manufacturing operators were drowning in data with no way to prioritize.
Zira's computer vision system generated thousands of anomaly signals per shift. Without a well-designed interface, operators couldn't distinguish critical failures from noise — which meant the technology's value was invisible to the people it was supposed to help.
How do you design for a world where every second of downtime costs money?
The question wasn't 'how do we display sensor data?' — it was 'what does an operator need to see in the first three seconds of looking at a screen?' Designing for high-stakes, high-velocity decisions required a fundamentally different approach than typical dashboard design.
I spent time on the factory floor before opening Figma.
I embedded with line operators to understand their actual workflow: how they scanned for issues, what caused them to act, and where current tooling fell short. That context shaped every information hierarchy decision in the dashboard — and it couldn't have been learned from a brief.
Design the interface where the data is born.
I contributed directly to the ML labeling pipeline and model evaluation workflow — which meant I understood the data at its source. That dual perspective let me design not just for what operators needed to see, but for what the system was actually capable of surfacing accurately.
A high-density dashboard built for the factory floor.
The resulting dashboard featured customizable alert hierarchies, camera feed integration, and at-a-glance anomaly summaries tested with operators in real production environments. The brand and marketing site I built alongside it positioned Zira as a credible enterprise product in a crowded industrial AI market.