The last two decades have seen a profound shift in the way most Americans work. Thanks to advances in communication and the move to an “always on” work model, the idea of an office has become broader, more open and collaborative. No longer designating a specific location, an office encompasses all the places where work is done. In the course of a single day, an office might include the passenger seat of a car or the counter of a coffee shop, a cubicle or a communal space, or even a traditional office.
In each of these situations, employees might leverage different communication channels on different devices to perform a given task—an app on a smartphone or tablet; a website on a laptop or desktop; or through a voice assistant, all the while expecting the same outcomes, regardless of channel.
The challenge for us at Capital One when building our commercial bank platform was to create a tool that responded to the new reality for the next generation CFO. This meant making sure that information is streamed to channels simultaneously. We also faced a second challenge – ensuring that, regardless of channel, users have the appropriate tools to view and manipulate this information.
In practice, we discovered that the benefits of this omni-channel platform go beyond accommodating a more fluid conception of the office. In addition to enabling people to work wherever they find themselves, the new platform also frees users to complete tasks on one device that were begun on another, and empowers people across an organization to collaborate. In effect, omni-channel platforms enable yet another breakthrough in the way people work.
Enabling a More Expansive Vision
Enabling omni-channel design is a series of digital toolsets called application programming interfaces (APIs), which software engineers build to create data conduits from one channel to another. For instance, customers can initiate an ACH transfer on the web, and the data will appear on the screen of any device being used by the accounts payable representative who needs to approve the transaction. APIs connect the various channels to a single source or transaction hub, allowing this flow of real-time data across channels and to precise destinations.
By harnessing the connective power of APIs, software engineers do more than facilitate the flow of financial information. Using APIs, our developers can quickly add new features to our platform, increasing productivity by enabling a more powerful, more convenient experience. For instance, after we launched our mobile capability, one of the things we consistently heard from our users was that they wanted better ways to access their information. APIs allowed us to integrate a natural language search product into our online platform, making Capital One one of the first commercial banks to offer natural language search on a digital platform.
APIs also enabled us to seamlessly integrate another third-party product – remote deposit capture. Capital One developers created APIs that allow our clients to capture remote deposits through our mobile app and process them using our dual control web application instantly.
The APIs we created for our digital platform don’t have to be limited for use in our own system. As part of our move toward open banking, we have created Capital One DevExchange (developer.capitalone.com), an online community that promotes the exchange of open source tools to help companies address such challenges as testing and debugging mobile apps and working more effectively in big data ecosystems.
As part of DevExchange, we are also posting a rapidly growing selection of our own APIs for our partners to integrate into their own apps and experiences. For instance, our Customer Transactions API, currently in beta, allows partners to integrate their Capital One credit card and bank account transactions data into their applications. We are also testing APIs that will enable partner apps to tap Capital One expertise for sign in and verification.
In the future, we could offer Commercial clients our APIs as a channel to enable them to access banking data from within their own treasury management or enterprise management system. This would give them the option of working entirely within their treasury management system, making it possible for them to accurately pinpoint balances and cash positions in real time.
Even more powerfully, our growing expertise with APIs gives Capital One the ability to think in more ambitious, more inclusive terms about our digital platform, gradually building it out to include credit information. For example, clients who have a real estate loan on an apartment could see the status of that loan and make a payment. Here again, APIs have the potential to foster a breakthrough in collaboration. By offering companies a comprehensive commercial bank platform that can pull data from one banking service to another, we can help them break down internal silos and gain a clearer, overall picture of their banking position.
Unleashing the Power of Machine Learning
The fundamental value of APIs is their ability to make pertinent data ubiquitous, no matter what the source, by interconnecting disparate systems. In the case of Capital One’s commercial banking platform, APIs increase productivity by enabling customers to complete a steadily increasing range of financial processes faster and more easily.
But APIs are just the first step toward an even more dramatic increase in productivity. As customers use our digital platform, they generate data that can be analyzed using the tools of machine learning to detect patterns in the way they work. Machine learning allows computers to learn without being explicitly programmed, and information generated by machine learning algorithms can be employed for a number of purposes. Most obviously, machine learning can be used to strengthen fraud detection, flagging variations from established routines for review. It can also be used to make it easier to carry out routine tasks. Having learned, for instance, that a customer routinely produces a set of overnight reports each morning, the platform can offer the customer the option of generating that report automatically.
Ultimately, machine learning can go beyond process simplification to process automation. In fact, with machine learning, the platform can be taught to perform entire back office processes, reserving employees for tasks that require judgement as well as experience. APIs greatly accelerate processes that require human intervention. Machine learning will enable the next great revolution in work by removing human beings from routine processes altogether.
In effect, technologies like APIs and machine learning that are at the heart of the commercial banking platform Capital One is building will do more than provide great experiences for our customers and support the way they work. These technologies will foster significant changes in the work itself, fostering collaboration, automating routine processes, and opening the way for dramatic increases in productivity.