Cloudification of Financial Simulations

Master thesis about High Performance Computing in finance using serverless and container based computing.

R&D Labs
R&D Labs
3 min readMar 29, 2018

--

Goal

In February 2016, I started writing my thesis at Ortec Finance master student in Computer Science. Ortec Finance wanted to investigate the possibilities of running compute intensive financial simulations in the cloud, with the use of the microservice architecture. Financial simulations are of vital importance for the company and developments in this area are thus watched closely.

The first day I got an introduction into the company, a laptop and spoke in more detail about the problem Ortec Finance was facing. Although a solution would have impact on many of Ortec Finance’s products, we decided to focus on one particular application. This was the risk analysis algorithm in the “Strategisch Beleidsinstrument”, that is built for Dutch housing associations.

My supervisors have always been clear that the main focus of my work would lie on my graduation and the academic value of my thesis; getting a solution to their problems was only a useful side effect of the research I would perform. A conclusion that techniques I was researching would not work, would be of equal value to them as the conclusion that those techniques would work. My supervisor at the university was very pleased with this mindset and maintains a close relationship for thesis supervision with Ortec Finance since.

Set-up

Generally, you’ll work full time on your thesis in the Rotterdam or Amsterdam office. By being physically present, you can informally ask help from someone else working at Ortec Finance outside the weekly meeting with your supervisors. In my experience, everyone is very helpful and interested in the application of you work. This is something I really liked about Ortec Finance.

The first few weeks I was investigating both cloud computing and microservices in literature. Eventually, we decided to focus on two cutting edge computing technologies at that time in the the working field: container based technologies and serverless computing, or Functions-as-a-Service. We chose two representatives of those technologies, namely Docker and AWS Lambda, and investigated the benefits and drawbacks of both technologies for our problem heavily. As both techniques were very new technology, still in heavy development, academic literature on both subjects was scarce. Ortec Finance provided me with all necessary equipment, such as access to cloud servers.

The people at Ortec Finance encourage creativity and new initiatives while writing your thesis, giving you much freedom in your work. However, help is always available and suggestions are continuously given to improve your work. I have enjoyed writing my thesis (well, most of the times…). The people and working environment made me decide to work at Ortec Finance after my graduation, and I hope to be working at Ortec Finance for many years to come.

Used tools & hardware: Amazon Lambda, Docker, NoSQL, Java

Author: Ferry

--

--

R&D Labs
R&D Labs

We work and experiment with both new modelling approaches and IT techniques and concepts in order to research their applicability to investment decision making