LIS Data Solutions – Demand forecasting and suggestion for procurement
LIS Data Solutions
Sector: Generalist
Business Case
Warehouse inventory management based on sales forecasting. Once the forecast is done, knowing the supplier’s LTsr, the level of service we want to deliver to the customer and the pending orders to be received, we come up with purchasing suggestions.
Objectives
Inventory optimisation Reduction of breakages Display and control of all items in the warehouse Exploitation of the BOM
Use case
Use of predictive algorithms: we have a system that brings together several predictive models of all types that allow us to adapt to different situations: based on the type of consumption of a catalogue item, we use some or other algorithms, always the one that best fits the circumstances.
Infrastructure
On Premise
Technology
Automatic or Deep Learning RPA
Data
We use structured data such as: purchases, sales, warehouse status, orders pending delivery, LTs from suppliers, etc. In this use case, we are talking about 25,000 different catalogue items in stock and active.
Resources
LIS human resources were 2 Data Science staff assigned for about 6 months. In the first project. We now have a lot of automated parts, which makes it much simpler. The client supported us during this time by answering data questions and defining the scope as well as the KPIs to be displayed. At the end of the project, everyone received training on how to use the tool.
Difficulties and learning
The main difficulty was how to make the predictions/suggestions from a business point of view. In the end, we were able to make sales predictions, break down sales into purchase products and make suggestions from there or, based on what is sold, we were able to break it down into purchase parts and generate predictions and suggestions from these data. We chose the first option after several months of discussions and tests.
KPIs (business impact and metrics of the model)
They currently have more than 25 KPIs in this tool. The impact on the business is very broad: previously they had a department of 3 people working for groups of suppliers. Now, one person manages purchases for the entire company once a week.
Funding
No
Collaborators, Partners
No
