Predictive and prescriptive analytics are concepts that are exciting for leaders, but hard to implement. Converge’s data science team knows how to apply advanced analytical models and spot opportunities to solve real data science problems. This case study describes how Converge helped a large financial services company use predictive models to better understand the factors that impact their business. Our client was ready to tap into their data and praised our team as the key to helping them successfully design, build and manage their own models.
The Challenge
Our featured client is one of the largest financial services companies in Canada. It has nearly 1,400 employees and prides itself for their diversified portfolio of asset classes and investments for clients around the globe.
Our client wanted to test and implement predictive analytics to see if there were factors that influenced or impacted their business volumes.
The goal was to deliver a predictive tool that could be adopted quickly by the team so they could begin to test their hypotheses. With a lot of great data sources to explore, but little experience with predictive analytics – they needed a guide to get this proof of concept off the ground.
Our Solution
The team had a wealth of knowledge in domains specific to finance, however, struggled with other data science tools in the past. Using Alteryx, we were able to explore datasets, create hypotheses, and test them very quickly, while allowing everyone on the team a chance to have input and get hands on experience with the tool.
This speed of iteration meant that the team was able to explore many possibilities in a short amount of time and resulted in a high quality model at the end. The tool’s ease of use meant that learning complex predictive concepts was easy, and being able to quickly try models helped solidify the predictive foundation of the team.
The Result
The team was able to use data to shape analysis and start forecasting trends and business volumes. By organizing and filtering the data, the client was able to unlock some of the value in previously underutilized datasets. Being able to forecast operational volumes allowed the business to better allocate both infrastructure and human resources, resulting in greater efficiency.
The final model allowed the team to run different scenarios so that they could plan for situations that they might not have encountered before. The knowledge they gained from this project opened the door to more predictive analytics projects and are currently in the process of planning new predictive use cases.