Banks Can Win The AI Talent War By Working With Fintechs

Subscribe to Our Blog

Artificial intelligence’s sweep through the financial sector is just underway, with serious landfall expected within the next 10 years. The smarter leaders are preparing for it. Earlier this year Goldman hired a top Amazon exec to build out an AI team. Not to be outdone, JPMorgan hired a senior Google exec and a leading AI researcher from Carnegie Mellon. Over the next two years, financial services firms are projected to invest as much $10 billion in artificial intelligence technologies – more than any other sector plans to spend.

Luring the big talent first is a necessary first step at closing what is a massive talent gap between the FANG companies vs. the rest of us. A third of firms surveyed by Teradata in 2017 said finding AI expertise is a major problem and a quarter said lack of talent (and adequate IT infrastructure) is getting in the way of broader AI implementation across the company.

The giant financial services companies can and do throw tremendous resources behind AI initiatives, but for everyone else, efforts are stifled by legacy systems, old-school data techniques, talent issues, and bureaucracy. A lot of companies we meet with have data scientists frustrated about spending more time on administrative and compliance-oriented tasks than on modeling challenges. Without clear direction, they find themselves working on expensive “science projects” rather than initiatives that can transform the company (and lead to bigger paychecks and promotions). And advances in data science are unfolding so quickly that it what their skills quickly atrophy if they aren’t applying the latest tools and technologies.

A recent report by the World Economic Forum on the future of AI technology in financial services underscores our experience. “Financial institutions”, it notes, “often lag in recruiting and retaining people with the knowledge, skills, and capabilities to create an AI-enabled workplace.” Among the talent and cultural challenges it highlights are inflexible employment relationships, a mismatch of skills and roles, a lack of ongoing training, and fiefdom issues.

So, what are some things that smart financial services companies can do to help bridge the talent and culture gap?

Look to build a great team – and then provide a clear path for career growth. The most important thing a business leader can do is hire great people – and keep them happy and engaged. Given the intense competition for data science talent, that’s harder to do than ever before. First, instead of focusing on hiring rock stars, look to assemble a great band whose members have lots of raw horsepower but are also eager to contribute to a broader team. Data science is changing so quickly that no single person can be expected to know or do everything. So where one employee has less knowledge and experience, another can support and help that employee grow. Make sure that you offer multiple ways for folks to evolve in their careers.

Break down bureaucratic structures and processes. To allow your data science team to flourish, you need to make sure that they are empowered to take risks and test out new and novel approaches. Our ZAML Explain tools automate a lot of model risk management compliance work so modelers can focus on the cool stuff. To be sure, this is not a call to relax controls or loosen compliance. It’s a call to explore ways to automate and innovate. For example, at Zest, we forgo department budgeting, org charts, job descriptions, and vacation timekeeping. Identify ways to do things – especially time-sucking and life-draining administrative tasks – smarter and more efficiently.

Foster a “learning culture” that balances research and applied science. Just as machine learning models are built to learn continuously, so are data scientists and machine learning engineers. Indeed, advances are occurring so rapidly that a significant portion of your data team’s knowledge will be outdated by the following year – or likely sooner. Our modeling team runs a book club-style group to help weed out good ideas from the lame ones. One team member highlights an article or new technique and presents it to the whole data science team. Another person tests it out on our “sandbox” of data sources and presents the findings. This way, the rest of the data science team isn’t getting sidetracked.

Consider bringing in outside expertise. For incumbents not wired up to attract and retain world-class tech talent, it can be helpful to develop a strategic partnership with a fintech firm that brings in a fresh perspective and focused expertise. Our experience suggests there are many ways these relationships can unfold. Some clients rely on us to take the place of an in-house modeling team, where our team of data scientists and programmers builds and maintain an end-to-end, AI-based underwriting solution. Others have relied on our software but also engaged us to create an internal training program for their own modelers once they realized much of their team’s knowledge was out of date. Whatever path is selected, it’s the combination of domain expertise and modeling expertise that delivers the best results.

Give us a shout if you want to work on building and deploying more profitable and fully explainable machine learning models in your underwriting business. We’re at