A New York based hedge fund with over $8 billion in assets needed visibility into the complex process of analyzing the historical performance of the trades made by their managers. They needed clarity and insight into a complex process, and needed to be able to analyze these trades based on dozens of different metrics. This analysis would take admins and managers hours every day to figure out, hours that could be better spent serving their clients directly.
Burrows Technology was hired to build a desktop application that allowed “quants” to design trading algorithms which could be dynamically plugged into the full application and allow traders to see charts and graphs on a dashboard interface. This application was written in C#, WPF, SQL Server, and used advanced techniques for dynamically loading assemblies into AppDomains. The project was started by in-house developers, and our initial engagement was for performance and quality analysis of the system. By analyzing database usage we were able to improve system performance over 100x.
Following this, we worked closely with in-house designers to improve the overall architecture of the system. We modularized the product to allow the discovery, creation, and loading of external algorithms that could be used for the trade performance analysis. This allowed the application to be modified without recompiling and re-distributing the application to all workstations. The assemblies were loaded from a central location, downloaded via HTTP, and loaded by the application while it was running.
This application was successfully designed and developed and ultimately turned over to the client’s in-house developers. The application was distributed to traders throughout the company and continually improved, algorithmically, by the in-house “quants”. This automated analysis saved hours of work every day for the traders and was able to assist them in analyzing opportunities for future trades.