Use Case

Nancy is the leader of restocking planning center. Her team main task is to overlook the restocking plan for 700 stores across the country. And the business is getting back on track after the restriction is lifted. But ever since, the number of stockout events was greatly increased, thus resulting in an increasing negative reaction from the client, as well as cross-store refill to cover the stockouts.
She needs a solution that help improving the demand forecast, and reducing cross store refill trip

To solve the problem, FPT data Science team joined hand with Nancy’s team, identifying potential features that can be used to let machine learning model learns and predict the trend better. Together, the team identified features from 3 major groups

• Business & public holiday calendar
• Google Trend
• Business “Sense”

With the improved demand forecast accuracy, the team went ahead and apply the Annealing Simulation algorithm to maximize the stockout reduction while constraining the inventory volume across all stores.
For 6 weeks period, by simulating the scenario where daily restocking operation would perform accordingly to the recommendation from the model, an impressive improvement in stockout and cross store refill (CSR) metrics was clearly shown.

Lan is a member of B2B sales team
Recently, the business is booming, everyday her team receive up to a thousand inquiries, far exceed the team capability

Mai is the leader of B2C planning center. Her main task is to support other member in introducing new plan to the user, to boost their usage. As the number of user is increasing fast, the team has new face joining every week. Thus, it is hard for her to cover keep up providing support to the team.
She needs an engine to automate the recommendation support, answering the question