Citigroup demonstrates the challenge of getting AI job in finance

So you're thinking of getting a job in artificial intelligence (AI) in financial services? Good luck with that. Even the big banks employ next to no one in their front office AI teams.

Speaking at today's AI Summit in London, Berkan Sesen, VP and global data analytics at Citi said his particular team comprises two people, or maybe three. - Three people in an organization that employs 231,000 people in total and 10,000+ in its institutional clients (ICG) group.

Sesen's small team is tasked specifically with the application of machine learning to marketing making, particularly for CDS trades. He joined Citi as an algorithmic trading quant in 2013, but his system has yet to be implemented. "It's still not in production but will happen soon," he told attendees this morning. "We are living in a transformational period in terms of machine intelligence."

And if you still want to get into AI? Expertise in speech recognition will help. Sesen said AI systems used in finance are increasingly reliant on the Markov model, a stochastic model typically employed in speech recognition programmes and used to represent randomly changing systems. "Hedge funds, systematic trading houses and banks all use Markov," Sesen said.

Sesen himself started out as a specialist in biomedicine, working for the NHS after completing a PhD in artificial intelligence and medical informatics at Oxford University. From there, he jumped in finance at Citi.

Contact: sbutcher@efinancialcareers.com