Resolving Order Dependency and Advancing Demand Forecasting for a Historic Seafood Processor: Yamasa Chikuwa Adopts No-Code Predictive AI 'UMWELT'

Tryfitting, an AI venture originating from Nagoya University, has announced the implementation of its no-code predictive AI platform, 'UMWELT,' at Yamasa Chikuwa Co., Ltd.

[Background] Dependency on Individual Experience and Inventory Disparities
At Yamasa Chikuwa, orders for mass-retail products were managed based on the experience and intuition of individual staff members. This led to challenges in standardizing workflows and handling staff turnovers. Furthermore, the high burden of daily order management required improved operational efficiency. Additionally, inventory levels varied significantly across stores, causing simultaneous issues of excess stock and shortages.

[Verification and Implementation]
During the pilot phase, inventory simulations were conducted for shipments to supermarkets in the Toyohashi area of Aichi Prefecture using UMWELT. After refining the algorithms, the company decided to proceed with full implementation due to promising results.

[Key Reasons for Adoption]
- Can be operated without coding, allowing for continuous onsite use.
- Flexible integration of the company’s complex order rules and workflows.
- Confirmed prediction accuracy and practical suitability during the trial.
- Received comprehensive hands-on support, including algorithm design and business process organization.

[Future Outlook]
While currently implemented in one department, the company plans to expand UMWELT across the entire organization, aiming to maximize sales and optimize inventory through demand forecasting.