Some of today’s best opportunities for banks to increase overall efficiency are found in streamlining operational processes involving documents, emails, and richer data sets with an intelligent automation approach. This can be especially effective in areas like customer lifecycle management (CLM), explains Kyle Hoback, director, intelligent automation, WorkFusion.
Whether responding to the pandemic or just recognizing a need to refresh their operating model, banks (like all major enterprises) have been increasingly relying on automation to increase efficiency.Â
What areas of banking are still ripe for transformation through automation? Some of the best opportunities focus on learning from data that is part of operational processes – documents, emails, and richer data sets – which can be used to train Artificial Intelligence (AI) through machine learning as part of an intelligent automation approach.
1. Streamline Processing of Documents From Banking Customers
Banks handle a huge number of documents, which customers share as part of Customer Lifecycle Management (CLM).Â
In commercial banking, customers opening an account must provide documents such as articles of incorporation, annual reports, certificates of good standing, and more. Consumer banking has some cross-over in the documentation but also includes other types like payslips, utility bills, bank statements, etc. Across all these materials, though, is a need to collect and verify relevant data elements, then enter them into core systems, such as Fenergo.Â
Automating how the institution handles “paperwork†is a key opportunity for banks today. Docs often arrive in various layouts and configurations, limiting the effectiveness of templates and rulesets. Because these rules-based approaches aren’t sufficient, operational teams of analysts have been manually processing the files, collecting and verifying data before updating the core systems. Even as banks have enhanced the customer experience with portals and chatbots, these upgraded user interfaces are often just a façade of a digital experience. The back-end updates have remained highly manual.
Intelligent Document Processing (IDP) leverages machine learning to understand data in documents and document-like data. IDP can read across formats and layouts to collect the necessary information as part of a larger process. People can still play a role alongside the automation, receiving escalations via Human-in-the-Loop (HitL). But with automation running most of the intake, collection, and entry, analysts focus more on validating and enriching the data sets and less on data-entry mouse clicks and keystrokes.Â
Account opening, with a highly manual back-end, for corporate customers may typically take 7–15 days and 1–3 days for consumer customers. With automation in place, the manual effort may be cut 70%, and onboarding time cut to 5–7 minutes. This leads to big savings: Organizations like Deutsche Bank are saving millions with these types of IDP solutionsOpens a new window .Â
Learn More: How to Supercharge Your Digital Transformation Journey With Connected Innovation Intelligence
2. Expedite Customer Email Servicing
Another key area for automation within banking involves frequent communication with customers via email.
Commercial banking areas such as Treasury ManagementOpens a new window often leverage email to keep relationships personable. Frequent requests like signer maintenance, statement copy requests, and address changes are delivered via Outlook email.
Without automation, banks typically resolve thousands of emails manually each month, creating a large operational drag. Account-facing personnel are often the targeted recipients, so this shifts their focus away from growing customer relationships and toward menial data-entry tasks like keying address changes into multiple systems.Â
Again, IDP can play a crucial role because email and attachments are really just additional types of documents. As with the document exchanges mentioned previously, this data is unstructured, precluding simple rules-based automation approaches. Successful automation must be smart enough to classify the intent of each incoming message, note whether the necessary information exists or is missing, then take the appropriate actions. If an inquiry contains sufficient data, the system may be able to rapidly fulfill the request in a near-real-time resolution. If not, the system can automatically escalate for resolution, perhaps by creating a ticket in ServiceNow or other tracking software.
Automation of email intake and fulfillment within an area like treasury management reduces 60–80% of manual handling time, with a vastly reduced processing queue, not to mention the renewed focus for account teams. Customers can continue to submit inquiries in a comfortable, full-text format yet benefit from more rapid fulfillment.Â
Learn More: What’s Next for DevOps Automation? 7 Predictions and Trends for the Year Ahead
3. Leverage and Learn With Richer Data Sets
Another way automation transforms banking is by allowing organizations to leverage and apply machine learning to richer, more expansive data sets.Â
For example, in Name Screening Alert ReviewOpens a new window , analysts must make countless decisions based on the information in an alert from software such as Bridger Insights. While an alert contains much of what’s needed for review, analysts also routinely perform supplemental research using internal systems, third-party sources like Dun & Bradstreet, and historical decisions from past alerts. This research makes for a lot of clicking and typing, adding minutes to a decision that would only take seconds if all the data was fully available.
Automation simplifies analyst reviews by pre-sourcing data before a manual review begins. Even better, automation can provide resolution without the need for an analyst’s attention. Because wider data sets are consolidated from all sources to support decisions (not just those contained within single systems), richer data can be used to continually train and improve through machine learning.
Automation for name screening alert review in this way typically reduces manual effort by 70% and reduces false-positive rates by 80%. Similar processes like payment screening alert review and KYC data sourcing achieve comparable results, all allowing analysts more time to focus on the most important areas.
Transforming Banking With Automation
As banks explore automation, they continuously look for ways to transform their operations. Automating with AI on documents, document-like data (such as emails), and richer data sets creates new opportunities for capturing value and strengthening banking operations.
Did you find this article helpful? Tell us what you think on LinkedInOpens a new window , TwitterOpens a new window , or FacebookOpens a new window . We’d be thrilled to hear from you.