A Fortune 500 personal care company wanted to bring better organization and order to its production planning. CPG manufacturers typically operate on stringent SLAs (Service Level Agreements) with their retail customers (Walmart, Target etc.) to ensure product deliveries are on-time and in-full (OTIF). Production planning and visibility about production capabilities across manufacturing assets and machines become imperative in such a case to limit delays in quantity and time.
MathCo partnered with the personal care company to help streamline their production planning. MathCo helped them design a robust end-to-end solution which not only improved the accuracy of the existing planning forecast models, but also supported real-time analysis through a user-friendly reporting framework.
A robust modelling and reporting framework that leverages tenets of next gen AI/ML techniques, was built and deployed on the Azure framework in the customer ecosystem, in place of traditional models with subpar accuracy. The deployed solution consisted of:
- Next-gen AI/ML model: This improved accuracy and helped identify potential delays in machine production. By leveraging variables of machine tenure, product parameters, historical performance etc. the model was able to predict the total time taken to produce cases of personal care products in a given production cycle.
- User-friendly UI: This enabled the planners to simulate and experiment with various scenarios on the fly allowing them to choose from the best options available in products and assets.
- Seamless integration with existing enterprise systems: Full-fledged deployment on the Azure system as well as API linking to SAP to extend workflow pipeline.
- Streamlined production planning processes and removed manual interventions and redundancies (blanket accuracy improvement ~ +5%).
- Estimated savings in fines due to mitigated delayed deliveries ~$13 million in the first year of operation.