On Monday Microsoft hosted the first “F# in Finance” at Microsoft’s corporate offices in London, co-organized by Fountainhead Consulting.
The London event was the first event of its kind and attendess including leading quantitative finance experts from major global banks. About half the audience were already using F# in production in financial enterprises, nearly all stayed through to the end of the event. Some uses of F# in production financial systems are described here: http://fsharp.org/testimonials
The event featured 10 talks including:
- Adam Mlocek (StatFactory) presenting on F# and Excel
- Stephen Channell (Cepheis) presenting on Using QuantLib from F#
- Michael Jones (SigmaFi) on F# for Managing Risk: a DSL risk language
- Daniel Egloff (QuantAlea) presenting “F# and GPUs: Faster, Faster, Faster”.
Daniel demonstrated F# GPU scripting from F# Interactive using Alea.cuBase based on F#’s unique meta-programming capabilities. This framework is already deployed in several major Swiss investment banks with some UK banks looking to adopt.
- Matthew Moloney (Tsunami) presented on “F#, Big Data and Machine Learning”
Matthew demonstrated big data programming with F# over Hadoop and Hive, using the F# Hadoop type provider.
- Yan Cui (GameSys) presented on “Scaling F# to the Cloud”
F# is used for some of highly popular server-side Facebook games at GameSys, a major games company in the UK with $5 billion turnover/year. The slides are here: http://www.slideshare.net/theburningmonk/f-in-the-cloud
- Jon Harrop (FlyingFrog Consulting) on “How F# Changes the Way We Work”.
Jon talked about how F# is now being adopted at a major insurance company. In particular he talked about how F# and DSLs have allowed much faster deployment of actuarial models into production. He emphasized that the actuarial programmers love F# and are very productive with it.
- Tomas Petricek (recently BlueMountain Capital) presented on “F# and Data Science with Deedle and R”.
Tomas talked about the F# and C# data frame library whose contributors include BlueMountain Capital. This is a high quality open library, giving F# a data frame library akin to the the Pandas library for Python.
- Don Syme presented on “Succeeding with Functional-First Programming in Finance”
Don discussed the rising importance of functional-first programming (Clojure, Scala, F#, etc.) and the characteristics of the significant number of successful projects being reported, for example at fsharp.org/testimonials. The business reasons for adopting functional-first techniques in business situations where tackling complexity generates revenue, especially for analytical programming, quantitative finance, insurance calculations, data programmability, GPU programming and testing.
Don demonstrated a new type provider for Xenomorph TimeScape financial data on Azure, and explained why the Visual F# tools from Microsoft are the safe and reliable choice for functional-first programming in the enterprise.
There were other talks too, you can see the full program here. Videos will be available in due course.
It was a great event, with most attendees staying right through to the last talk.