THETRADETECH DAILY
SPONSORED ARTICLE
Patrick Flannery this technology. They ' re going to have to find a way to buy it.
What is the most promising path to success in this development? To date, the success rate has been pretty small. The hyperscale technology companies have the resources, the pipeline, and the people that can build these solutions. Within capital markets, it ' s the most technically capable firms. Instead of looking for client facing products, firms should think about their own internal use cases, looking at the quality, the transformation, the integration of data and writing code for themselves. One of the reasons we focus so much on data quality and coverage is because having access to high-quality and trustworthy data in the same format that can underpin AI models is of critical importance. Having exchange data with the highest fidelity to what the exchange puts on the wire enables AI models to use our data effectively and efficiently.
Another issue is evolving legacy systems, given there ' s a lot of old code embedded in them. Using AI to rewrite these systems is something I ' ve heard about, and I haven ' t seen it be successful at scale yet, but I ' m particularly interested in it. It could make businesses dramatically more productive through lower costs and more agility. They would also have the ability to be less married to any one existing solution.
There ' s plenty of new models and activity. It’ s difficult to declare success or failure. Companies are coming out with things one after the other. They’ re improving and they ' re becoming more efficient. These models allow firms to be better trained, more flexible, and more efficient. The key thing now is to take those and say ok, how do we use it?
www. thetradenews. com 7