Credit Suisse Case Study
Enabling Scale and Performance
- Estafet Team
The Story
At Credit Suisse, Estafet tackled critical challenges to improve system resilience, scalability, and performance. The team addressed frequent disruptions caused by core risk systems, redesigned platforms to streamline scaling, and separated UI queries from batch feeds to optimise processing efficiency. By enhancing load balancing to prioritise feeds by type, locality, and size, Estafet ensured that vital data reached the business without delay. A comprehensive performance review revealed that the system was nearing the limit of vertical scaling, leading to recommendations for horizontal scaling, fault tolerance, and best practices in platform configuration. These efforts resulted in a more resilient infrastructure, faster data processing, and increased scalability, supporting Credit Suisse in maintaining timely access to essential market risk information. Estafet’s contributions provided insight into existing designs, offered rapid problem diagnosis, and introduced new resilient design implementations, ultimately strengthening Credit Suisse’s ability to handle future demands effectively.
Challenges
- RESILIENCE OF DATA FEEDS
- SCALABILITY AND AGILITY OF RISK PLATFORM
- EFFICIENT PROCESSING AND PRIORITISATION OF DATA
Optimising Information Priorities
Prioritising the right information was critical to supporting Credit Suisse decision-making. Estafet tackled the challenge of inefficient load balancing, where incoming feeds were distributed across multiple queues without considering the priority, locality, or size of each feed. This led to delays, as long-running or less important tasks would sometimes block the processing of more critical feeds. Estafet’s solution was to implement a smart prioritisation strategy, focusing on feed type, size, and importance. By prioritising feeds based on business relevance, they ensured that key data was processed first, reducing bottlenecks and idle time in the system. This approach not only optimised throughput but also improved the timeliness of the most critical information, enabling Credit Suisse to make more informed, faster business decisions. The new system allowed for dynamic resource allocation that adjusted based on business priorities, ensuring the most important data always received the attention it needed, which ultimately enhanced operational efficiency and supported better decision-making.
The Solution
Estafet provided a comprehensive solution to address Credit Suisse’s key challenges in resilience, scalability, and performance. The team rebuilt the PowerMart worker software to ensure continuous, resilient data feeds, eliminating connection drop issues and improving throughput. For WebLogic, Estafet redesigned resource allocation and developed a scalable model, enabling more efficient scaling and reducing reliance on costly vertical scaling. To optimise data processing, Estafet separated UI queries from batch feeds, ensuring that each received the appropriate resources based on priority, which resulted in consistent and faster feed processing. They also enhanced load balancing by prioritising feeds based on type, locality, and size, ensuring critical data was processed first. Finally, Estafet conducted a performance review of WebLogic and provided recommendations for horizontal scaling, fault tolerance, and best practices in configuration, testing, and clustering. This holistic approach improved system resilience, agility, and scalability, enabling Credit Suisse to meet increasing business demands.
Deliverables
- RESILIENT SOFTWARE DESIGN AND REBUILD
- PLATFORM SCALING
- BALANCING LOAD AND PRIORITY
The Success
Estafet’s impact was to resolve critical infrastructure challenges to improve resilience, scalability, and performance. This included the rebuild of core services to ensure continuous data feeds, eliminating connection drops that caused delays in delivering vital business information. Re-engineering the platform also meant that Credit Suisse could scale out without having to buy more powerful (i.e. expensive) machines. Finally, the separation of (heavily computational) UI queries from (business critical) batch feeds meant that Traders received important, relevant data quickly and consistently so that they could make better decisions. These strategic improvements led to a more resilient, efficient system, enabling Credit Suisse to provide timely market risk information and stay competitive in a fast-moving financial environment.
Outcomes
- IMPROVED DATA TIMELINESS AND RELIABILITY
- ENHANCED SCALABILITY AND COST EFFICIENCY
- OPTIMISED DATA PROCESSING AND THROUGHPUT
