AI & Data Science in Trading London: Commentary & Thoughts

Anthony Tassone

September 27, 2019

The recent AI & Data Science in Trading conference in London was an excellent opportunity to meet some of the incredible data scientists in financial services who are working to advance technology across trading disciplines, from pre- and post-trade analytics to data-derived alpha generation. I was pleased to attend and represent GK, in particular as chair of the Natural Language Processing stream on Sept. 16.

The NLP stream featured a panel discussing the challenges and opportunities to be considered in widespread NLP adoption from both vendor and user perspectives. The panelists and I agreed that NLP is rapidly becoming widely accepted and used in many aspects of trading and banking, and the next few years will surely bring new frontiers in its capabilities as barriers to adoption including privacy concerns, compliance and legacy technology integration are solved.  

Indeed, a common theme in my conversations with conference attendees throughout the two-day event was the insights to be found using NLP, especially in solving workflow challenges. Financial institutions are looking to collect and analyze sample reports in three main areas:

  1. Customer insights including methods of communication and sales touchpoints
  2. Customer liquidity metrics including the market impact of each customer trade, both immediately and in the longer term
  3. Internal sales performance analysis including for each trader and product

Along with providing validation for GK’s products, these conversations are critical as we continue to hone our technical functionalities to meet client demands.

The conference also featured many academics who brought interesting viewpoints and research for thought, as well as industry science powerhouses with technical talks for quant and research-focused attendees.

I really enjoyed the presentations by Gary Kazanstev, Head of the Machine Learning Engineering Team at Bloomberg, who gave a technical presentation on the recent advancements in machine learning. Gary discussed the huge amount of problems that Bloomberg is trying to solve with machine learning. 

Overall, GK appreciated the opportunity to be a part of the scientific community represented at this event, with special thank yous due to the conference organizers and my colleagues in the NLP stream.