To ensure high AI performance over time, AI teams need to have a performance monitoring system for their deployed machine learning models, and trigger model re-training when performance degradation or drift exceeds a threshold.
Suntory understands this to be critical when deploying their logistics ML models to predict pallet usage and recovery in their distribution channels, which has a significant impact on profitability. Suntory chose Citadel Radar and Citadel Lens as their platform for ensuring reliability.
After using Citadel AI’s products, Suntory was able to improve their model accuracy by detecting the timing of required model re-trainings, preventing bad training runs that lead to performance degradation, and automating operational tasks such as model version management and drift detection.
Suntory plans to apply Citadel Radar and Citadel Lens across a variety of AI models in the future.
For more information, please visit Suntory System Technology’s website to download the case study (Japanese only).
About Citadel AI Inc.
Citadel AI provides automated testing and monitoring products for AI applications, to help organizations prevent technical and business risks from AI bias, and maximize AI performance from PoC to operation.
Citadel AI’s co-founder was the former product manager for ML infrastructure at Google Brain, and is currently leading Citadel AI’s product and engineering teams.