Profinit

PROJECT SUMMARY: Profinit built a custom-made big data computational platform, based on the Hadoop, Apache Spark and Python technological stack. Profinit delivered consulting, expert and development services. CLIENT'S INTRODUCTION: Ceska sporitelna is the largest bank in the Czech Republic with 4.5 million clients. My position is Tribe lead responsible for data management and analytics. PROJECT DETAILS: We decided for Profinit due to demonstrated expertise in Hadoop technologies and positive experience with Profinit services in other projects within Ceska sporitelna (e.g. application development). Profinit proposed a team of experts (5 people) with the expected background and expertise. The project aimed to prepare data integration for the source of payment transactions, collect all relevant data in a Hadoop-based data mart and build an analytical model based on those data to identify fraud or suspicious transactions. In the next phase, those models were integrated into an online transaction monitoring front-end application. Profinit helped us to set up overall governance and analytical model lifecycle management. RESULTS & FEEDBACK: Thanks to project deliveries we realized a significant decrease in fraud transactions and improved clients' NPS ratio. The workflow between our team and Profinit's team was absolutely effective. Profinit dispose with the culture of self-managed teams led by senior consultants or architects, optimal mix of senior and junior people. We reached a high level of efficiency (we realized cost savings in projects in comparison to the original estimation). Their services were also highly cost-effective.