Case Study: Fleet Management Machine Learning Solution for a More Efficient and Less Manual Process
One of Altron Karabina’s clients, a vehicle fleet management company, experienced significant challenges with their manual process for predicting residual values (RV) of fleet vehicles before preparing leasing quotes. This was vitally important to their business – if they couldn’t calculate the expected value of the asset after the lease, then they wouldn’t be able to calculate the correct monthly payments to ensure the required margins.
There were only a few people in the organisation who could work with the complex pricing tables and it took them a long time to come up with a verified result, which caused significant bottle necks in the quoting process.
The Solution – Fleet Management Machine Learning
After understanding the client requirements and their business challenges, Altron Karabina suggested implementing awhich would use historic vehicle sale prices instead of the vehicle pricing tables to calculate a price after the lease with an accuracy of 94%.
The technology framework used Azure Databricks, Python scripts, SQL Server 2017, PowerApps, andfor the presentation and interaction layer.
All of the data was visible at a glance through Power BI and all processes were automated. The solution was thoroughly tested against the manual results and performed well within the required parameters.
Altron Karabina enabled the client to run their process much more efficiently, in a fraction of the time, with far fewer manual interventions whilst achieving the same results.
The fleet management machine learning solution was delivered within a 5-week time period – on time and within budget.
“I’d like to take this opportunity to congratulate the Altron Karabina Team for doing such a great job on the Residual Value tool. The expertise and level of professionalism shown by all was really refreshing and we achieved what was promised.”
if you would like to meet about similar solutions for your company.