The all-digital bank has teamed up with Microsoft to explore the possibilities of quantum computing, an emerging technology for which talent is scarce and tools remain immature.
By Clint Boulton
Sakkmesterke / Getty Images
Quantum computing is much ballyhooed for its potential to run simulations vast amounts of data, ideally to help enterprises solve hairy optimization problems. And while the relative immaturity of quantum technology has dampened some of its attendant enthusiasm, one financial service provider is forging ahead.
Ally Financial has partnered with Microsoft to leverage the tech giant’s quantum computing assets in its Azure cloud. With quantum, Ally is aiming to run sophisticated simulations exploring new financial products for customers more efficiently than with conventional computers, says Sathish Muthukrishnan, the all-digital bank’s chief information, data and digital officer.
“We believe quantum is well suited to solve a range of complex computational problems,” Muthukrishnan says. Ally must “constantly innovate and look at what’s next” in a competitive environment, he adds.
One way to grasp the differences between conventional and quantum computing is to consider how each might explore the optimal route to the end of a vast maze, says Muthukrishnan.
In attempting to get through the maze, the conventional computer methodically works through its binary bits, hitting one dead end after another until it reaches the end. Conversely, the superposition property enables a quantum computer to travel all maze paths in parallel until it reaches the end. In summary, quantum solves problems by considering many more variables while arriving at answers in much shorter times than conventional computing, Muthukrishnan says.
Partnerships are critical for clearing quantum’s hurdles
To tackle the nontrivial challenges quantum poses, Ally joined Microsoft’s Enterprise Acceleration Program to explore use cases, leveraging software languages, APIs, and infrastructure to build quantum skills. Microsoft’s Azure Quantum cloud software is more than capable of processing information at high velocity and the lowest margin of error, according to Muthukrishnan, adding that the software maker operates with the speed of a startup but is backed by the vast resources of the enterprise mothership.
Ally’s business model is akin to Muthukrishnan’s maze analogy. With its variety of financial products, including anything from mortgage and auto loans to corporate lending and vehicle insurance offered across its core B2B, B2C, and B2B2C channels, Ally has a seemingly infinite number of permutations to investigate for business value.
Quantum can help scale Ally’s artificial intelligence (AI) and machine learning (ML) models to solve the optimization challenges confronting Ally’s business model, says Marcos Souza, Ally’s executive director, head of AI and advanced analytics.
One avenue Ally is exploring includes churning through large amounts of market data, including historical and geographical information, to determine the best prices for financial products. Another optimization challenge includes how to remarket leased motor vehicles at a time when the market is hot for used vehicles, an effort that hinges on multiple variables, including supply and demand, pricing, and geography.
“Quantum allows us to run same things on traditional AI and ML models but with a much faster highway to get results faster,” Souza says.
Ultimately, Ally is exploring quantum to land on the right product for the right customer at the right time and at the right price, says Muthukrishnan. This aligns with Ally’s mission of using technology to treat each customer as its only customer, what Muthukrishnan describes as Ally’s “segment of one” philosophy.
While COVID-19 prevents employees from either company to work together in person, Ally and Microsoft engineers regularly participate in digital whiteboarding sessions to brainstorm ways to optimize Ally’s business.
Frequent collaboration aside, the challenges are considerable. Investing time, talent, and money in quantum requires Ally to carefully consider the right optimization problems to prioritize and then build out the “muscle memory” while exhibiting the “enterprise patience” to pursue those projects, Muthukrishnan notes.
Many financial services firms, including Barclays, JPMorgan, and Bank of America, are exploring similar use cases for quantum. But progress has been slow due to technical limitations and the scarcity of talent to work with it. As Ally notes, quantum hardware is not yet reliable enough to scale for large production workloads.
“It’s going to take time,” Muthukrishnan says. “But as the tech leader for a digitally native company, my job is to see the next big thing around the corner.”
If there is such a thing as an early mover advantage in quantum, Ally is on the right side of the timeline, as Gartner estimates that mainstream applications for quantum are still more than a decade away. The researcher expects nearly 40% of large enterprises will devise quantum initiatives to build management skills ahead of quantum computing opportunities by 2025.