Zafin Hackathon: Machine Learning and AI Algorithm

Machine learning hackathon

With contributions by: Rajnish Ranjan & Shubham Jaiswal

One of the great things about being a tech company is the ability to concentrate innovation in the context of a hackathon. In essence, a hackathon represents a small time frame where our programmers work together to collaborate and create something incredible.

Zafin organised its annual “Machine Learning Hackathon” on March 8th and 9th to build robust, efficient and effective machine learning algorithms for prediction of cross-sells and upsells of banking products to existing clients. Our other focus was to predict churn through a “customer-centric” approach. By focusing on the needs of the customer, our algorithm could act autonomously to ensure they were met in the quickest way possible.

We split our programmers into teams, to encourage a bit of friendly competition. Who could code The overall objective of our hackathon was to build a best in class predictive algorithm to accomplish the above goals — and to do it in only 48 hours. No small feat!

Collaborative Competition on Machine Learing

The participants were presented with a North American bank’s data, and were required to build machine learning algorithms to identify the right customer for a right product at the right time. This would help maximize non-interest income for the bank, and preemptively identify potential churn customers, giving banks a chance to improve their customer experience and keep their business.

In spite of the time crunch, our teams did great! They managed to build a robust, efficient and effective machine learning algorithm by mining features based on customer demographics, behavioural pattern, psychographics, transaction pattern and payment preferences.

But there could only be one winner. Within the carefully defined timeline, the winning team displayed their agility and statistics skills by performing data engineering, pattern analysis, sequence mining, and much more. These kind of events helps us to be best in class and allow our strong team of data scientists to prove their mettle.

Zafin’s Chief Analytics Officer Suman Singh says:

“Zafin’s vision is to apply AI and advance machine learning to empower banks in reaching out to their customers for easy banking. Currently, banks have been faced with issues in a complex IT environment and has been unable to harness the power of big data and deep learning. Zafin is committed to help their clients to quickly build, implement, monitor and maintain advanced machine learning algorithms with a single click through their analytics platform “Zafin Insights” in cost effective manner. Zafin Insights offer seamless integration with any core banking legacy system and establish a common operational framework across all IT platforms enabling banks to use predictive scores and insights in making effective decisions.”

This year’s Machine Learning Hackathon was a huge success, and we’re incredibly proud of our team’s dedication to innovation, and their ability to create cutting-edge solutions to give our clients a competitive edge.


About Zafin

Zafin (@zafin) is a leading financial technology provider that enables banks to form richer, more personalized client relationships. Built from the ground up for financial services, its platform empowers banks to enhance revenue and operational efficiency. Founded in 2002, Zafin sits among North America’s top FinTech companies, and is trusted by retail and corporate units at some of the largest banks worldwide. Headquartered in Toronto with global offices, Zafin has a proven track record with a 100 percent client retention rate as validation.