Dynamic pricing for more efficient workflow
The need to create quotes individually was slowing down our customer's work processes. With an automated pricing system based on data science and machine learning, we were able to sustainably reduce the administrative effort.
Task: Less work involved in creating individual quotations
The preparation of quotations was enormously time-consuming for our customer. Each inquiry was made individually, which meant that the offers also had to be calculated, written and sent individually each time. With hundreds of inquiries and potential leads, this was an effort that should not be underestimated and that distracted the employees from the actual company goal: to open up new markets and generate more sales.
Our goal was to use dynamic pricing based on data science and machine learning to reduce the amount of work involved in quoting and thus provide a better experience for the end customer.
Solution: Automated pricing with data science and machine learning
In collaboration between Data Scientists, Machine Learning Engineers and our MLOps team, we developed an automated pricing system that captured the customer's previous quotes and used them to calculate further prices.
The results are available accurately and quickly, so that prospects benefit from a higher level of service and receive an immediate basis for decision-making for their order. For our customer, in turn, automation means enormous time savings in lead generation and thus a more efficient workflow.
Result: Improved work processes and opening of new market segments
For the company, this was the first data science business service to be provided via the cloud. We were able to score points primarily due to the rapid implementation: It took just two months from project start to customer onboarding.
Our model demonstrates both the feasibility and meaningfulness of data science and machine learning for the customer. More extensive integration is already being negotiated - to open up new market segments, but also for better internal collaboration.
Goals achieved
Design, implement, secure, and deploy a cloud service (API and web application) that can be used by third-party providers
Responsibility for regular maintenance, customer support and onboarding
Setting up development processes and automation for Data Scientists
Design model wrapping and MLOps processes to deploy and reuse the pricing model in on-the-premises services.
Establish real-time monitoring dashboards and data science feedback loops to track and improve model KPIs