Is there a growing demand for artificial intelligence-driven functionality from clients?
The interest in AI has exploded in the past few months, crossing over to capture the imagination of both consumer and business markets. However, rather than a growing demand for artificial intelligence-driven functionality per se, I would say it’s part of the ongoing keen interest in any new technology innovation that can alleviate operational friction and improve how our clients interact with their technology. They aren’t necessarily coming to us to say, “we want an AI solution,” but instead, asking us to deliver ways – through the most appropriate technology – to streamline workflow further.
Although AI has progressed dramatically, concrete use cases are still somewhat in their infancy, meaning that while the technology is accessible, applications are still being formulated. As a technology partner, it’s our job to innovate beyond expectation and, essentially, stay abreast of these technological advances and ensure we apply them in the right manner. For example, to streamline a point of friction that a trader might not realise could be optimised or refined previously.
ChatGPT may appear to be a recent development, however as part of our research & development team efforts, we’ve been working with OpenAI Codex API for some time, having signed up for the early beta. It gave us the opportunity to get ahead of the curve when understanding and refining its initial various use cases. The outcome is our recent innovation FlexA.
We’ve been showcasing the solution to clients and prospects during our regular discussions. Presenting real use cases means that clients can immediately see the value of applying the technology, which in turn is now driving the interest and demand – and crucially, they are feeding back ideas and suggestions to take things further, which is what industry innovation and driving change is all about.
How can artificial intelligence revolutionise how trading teams interact with their data and trading technology solutions?
Our initial work with AI has seen us focus on two key areas – reducing the friction between traders and technology and increasing the speed at which traders can work with, and understand, vast volumes of data.
Taking working with technology as a starter, we’ve created a new AI layer using ChatGPT, which sits across FlexTRADER EMS and can simplify the interaction between the trader and their trading technology platform. Doing so enables a trader using the solution to use natural language to control their FlexTRADER user experience via voice or chat box, with AI acting as a translation layer between human and machine.
To provide context, traders can use voice commands to quickly establish a deep situational understanding of the order blotter via filtering, sorting and surfacing the data they need to see at any moment. Simple requests can be made, for example, “how many of my orders have been waiting for more than ten minutes?” or “give me a breakdown of value by portfolio manager”. Taking a step further, orders can be grabbed to prep or allocated internally across the desk. For instance, “assign all energy orders” to a specific trader or “send all orders under 1% ADV to the GS VWAP”.
The second area where we see a substantial uplift is working with vast volumes of trade history and TCA data. Using AI, traders can request ad-hoc reports with specific visualisation formats, including tables and charts, to assist with pre-trade decision-making. It removes the need to formulate complex search queries. Instead, it enables traders to use natural language to rapidly interact, interrogate and understand data in the manner that makes the most sense to them. For example: ” what was the most I’ve paid for Apple in the last year? ” or “show me a chart of brokers in July by value”.
The important point is that using AI doesn’t replace the trader, but augments to speed up and improve productivity. For example, while AI can surface and visualise data – and reduce the number of clicks or customed queries required to get to it, it still needs analysis, understanding and expertise to action it. Further, while it can organise and navigate an order blotter, the trader is still in control via safeguards to send the order.
What other technological advancement are you expecting to see in the EMS space?
While understandably AI and ChatGPT might be dominating the headlines due to its groundbreaking nature, accessibility, and unlimited potential uses, an area we see advancements in – across asset classes – is the drive to use the EMS to tackle an increasingly fragmented liquidity landscape.
To address the dearth of accessible liquidity prevalent across all asset classes, the EMS is becoming a key component to effectively democratise liquidity by directly streaming sources of traditional and non-traditional actionable liquidity into the order blotter.
For example, on the equities side, we see requests for seamless integrations to non-traditional sources of liquidity, such as SI and CRBs, and to have this available alongside more traditional offerings. From a fixed-income perspective, bilateral electronic connections are finally becoming commonplace.
Not only does having both traditional and non-traditional sources of liquidity deliver a holistic view of options available within the market, but it can also mean that automation rules can be applied, further streamlining workflow. As firms increasingly look to stream and democratise liquidity data directly into the blotter, cutting-edge EMS technology is the key to handling the vast volumes of analytics from bilateral connections.