“And it won’t stop here, because Luxembourg’s goal is to become one of the most advanced nations in the data-driven economy.”
This comment comes from Inès Buttet of Luxembourg For Finance, as part of her opening remarks at the LFF’s “Focus on AI & Big Data” event on Wednesday morning.
The “it” she refers to is artificial intelligence and particularly its usage in the financial industry. Well beyond its early stages of being able to find patterns within datasets, AI can now, according to Buttet, produce increasingly reliable output.
What kind of output and in what kind of applications? For some, chatbots come first to mind. “It’s really about time that this view changes,” says Andreas Braun, director of AI and data science at PwC Luxembourg, of the fact that chatbots remain the status quo for AI usage in the financial sector. “AI is all around us, and it will shape our world to be for the next coming years.”
In that spirit, here are six use cases beyond chatbots for AI in the financial sector.
Use case: decision support
Plenty of software already exists to help professionals make simulations, predictions and buying/selling decisions, says Braun (PwC). “But have you thought about the executive sphere?” he asks. “Right now you’re doing your strategy for AI… have you thought about using AI to really create your strategy?” Applications facilitating this type of shift are only months away, he adds.
Use case: hiring
Braun (PwC) points out that recruitment is tough in the financial sector, with competition fierce. Could AI be a way of approaching this issue? Potentially, he says, though adds that this would fall under the high-risk category of new AI applications. “When you’re filtering [your CVs] autonomously, with the help of AI… you’re kind of taking a decision on the future career of this person,” he says. “These applications will have to be monitored in the future.”
Use case: sentiment analysis
“You can [with natural language processing applications] increase your sentiment analysis,” says Frédéric Moioli, managing director of Lingua Custodia, “and detect weak signals from your customers.” Such NLP apps can scan social media accounts, he says, to build a picture of what’s going on in a company’s particular marketplace, enabling it to adapt quickly.
Use case: data management
NLP models are better than humans at some tasks, including extracting value from large bodies of text, says Davide Martucci, CEO of Next Gate Tech. From there, classification algorithms can be used to map, harmonise, clean and enrich data, which would be valuable in particular regarding data in its early or raw stages, according to Martucci (who is speaking in a context of fund industry operation processes, though ostensibly the principles could apply elsewhere).
Use case: contextualising data
The other use case Martucci (Next Gate Tech) highlights is at the end of the data-flow: the creation of analytics. “Different models of machine learning can be implemented at different times in the value chain to do different tasks,” he says. This would pertain to the contextualisation of events, he says, useful for anyone doing oversight or calculating risk. “Let’s say you have an oversight solution that performs ESG score and ranking. [The benefit would be] being able--thanks to artificial intelligence--to actually contextualise those different elements, giving the final user even more clues and indications of what the underlying cause is.”
Use case: translations
There is a huge private bank, says Moioli, that has done a lot of research and mobilised tons of data in order to build a profile of a customer. This bank has offices in the UK, Luxembourg, China and Hong Kong. “They’re doing the research in Hong Kong… then they translate it into English, in a minute,” he says. “And then the results could be used… in Luxembourg or in the UK at the same time as in Hong Kong. So it helps [the bank] be more proactive.”