Izy AS set out to modernize workplace canteens with mAIfood, an AI-driven checkout and prediction platform designed to reduce food waste, eliminate queues, and bring data intelligence to daily operations.
To make this vision commercially viable, Izy needed a technology partner capable of building a high-speed, reliable, and scalable computer vision system that could operate under real-world constraints.
That partner was Resimator.
X
Higher Throughput
%
Lower Operating Costs
Ultra-low latency requirements: Checkout needed to be near-instant to avoid queues and lost throughput during peak periods.
Hardware + software integration: The system had to accurately recognize food items using cameras and scales, while coordinating with payment providers and logistics services.
Unpredictable operating conditions: Hardware components, third-party APIs, and product catalogs were evolving in parallel with development.
Scalability across multiple sites: The solution needed centralized analytics, role-based management, and consistent performance across canteens
Low-latency computer vision pipeline: High-resolution image recognition combined with integrated weight validation to instantly identify food items on a tray.
Fault-tolerant backend orchestration: Intelligent resource allocation ensured uninterrupted service even when hardware or third-party services were under development.
Scalable, role-based system design: Centralized management and analytics made multi-site expansion frictionless. The result was not just a faster checkout, but a data-driven operational platform.
4× Higher Throughput Without New Infrastructure (Revenue, Speed): Ultra-fast checkout eliminated bottlenecks, enabling significantly higher transaction volume during peak hours using existing space and hardware.
Reduced Dependency on On-Site Staff (Cost): Autonomous self-service lowered checkout personnel requirements by 15%, improving labor allocation and margins.
Enterprise-Grade Reliability (Risk): A 99% service reliability rate protected revenue by minimizing downtime and operational disruption.
Data That Improves Decisions Over Time (Revenue, Cost): Continuous learning and analytics turned checkout data into insights for demand forecasting, inventory optimization, and waste reduction.
