Bloomberg utilizes functional programming (OCaml) in production for powering part of its Bloomberg Professional Services product. Aided by a collaboration with LexiFi, several Bloomberg applications are now making use of GUI generation and financial contract representation; Bloomberg furthermore has embraced OCaml to develop domain specific languages and the problem of integrating the vast amount of data available in the Bloomberg infrastructure.
This talk will detail our experience of the adoption process of functional programming by a team of primarily C++ developers and how we plan to extend the use of OCaml to more applications. We will use the concrete business problem of trade capture to illustrate our technology transition.
New financial products are created continuously and the ability to quickly develop and deploy applications to capture the details of such contracts accurately is crucial. While one can easily build generic screen to record unstructured data, difficulties arise when complex validation and UI behaviors need to be implemented. Bloomberg has been through many iterations in the development of solutions to this problem. At one time, code generation techniques were employed; this technology was then migrated to benefit from the modified OCaml compiler developed by LexiFi (MLFi). This technology is now fully integrated into the Bloomberg development environment and makes use of all available proprietary APIs.