Artificial intelligence and automation are quickly moving beyond being buzzwords and becoming an integral part of the digital transformation of our world. Banks too need to deploy these technologies at scale to remain relevant. AI and Automation has the potential to drastically transform front to back-office operations using a next generation workforce.
AI can help augment various human capabilities such as vision (read documents, images, etc.), listening (understand natural language and arrive at the intent), decisions (suggest decisions based on continuous learning), actions (understand and execute instructions), sense (sense and respond/adapt) and voice (recognize speech and voice). Using these capabilities, each banking process that involves bank staff can reviewed and re-wired to create “human ++” capabilities.
Currently, improving customer experience has been the primary focus for most banks’ AI implementation, and that is certainly an important area to address. However, expanding the focus to empower employees through AI and automation is equally important. This can help banks accrue greater benefits. Bank staff do spend considerable amount of time with the systems. The nature of task include mix of cognitive, repetitive, and trivial
Therefore, prioritising employee experience and providing them with the right tools to augment their capability is a good strategy. Tools that offer hyper automation and intervention by exception, and are flexible and user-friendly can help provide a frictionless experience of bank staff and translate into greater productivity. It can not only help improve process efficiency, but it also brings in greater effectiveness, thereby ensuring happier employees, but also happier customers. In addition, it can also support by making simple decisions independently and supporting complex decision-making.
Let’s take the example of Al Ahli Bank of Kuwait (ABK), which is among Kuwait’s leading banks. When the bank was using manual salary processing for its small business customers, the process would take up to a week. Apart from the sheer delay, the process was riddled with error and dependencies. With automation, the salary details provided by customers are digitized using a QR code and processed with the bank’s core banking platform. The solution executes 95% of requests and assigns only the remaining 5% exceptional cases to staff.
Not only has there been a 92% increase in productivity due to elimination of manual effort, but the time needed to process a request is down by an astounding 97%, allowing salaries to be credited into customer accounts within six minutes. Other gains include improvement in accuracy and high scalability to handle month end peak volumes with ease.
Augmenting employee capabilities at each stage
Irrespective of their job role, each bank employee’s tasks typically consist of a mix of some cognitive work and some trivial and repetitive tasks. Let’s break down the typical processes in banking into some generic steps and see how AI and automation can help augment each step.
Getting input
The input process – whether it is extracting information from scanned images , documents or reading emails and deciphering the intent or listening to a voice call etc., largely involves trivial work that can be performed via AI applications. For instance, AI can read documents, images, emails, voice inputs (listening to a phone call), process structured and unstructured inputs and translate them into structured inputs.
Enrichment
Before additional processing, bank staff requires a lot of augmented information to enrich the input data. Enrichment tasks, for example, include extracting profile information, gathering historical information etc. These tasks as well as others such as identifying patterns to determine Anti-Money Laundering, suggesting new products or services, and gathering market inputs relevant for decision making can be automated.
Decision making
While decision making involves cognitive skills, AI can certainly help augment the process. For simpler decisions, AI can be leveraged to make independent decisions without human intervention and approval. For more complex decisions, AI can make recommendations and leave the final review and approval to the employee. For example, AI can recommend if a certain loan application should be approved or rejected. It can suggest relevant new product or service options that can be marketed to a given customer. It can even provide cash flow projections for an organization based on historical performance and market data. In addition, AI can help highlight transactions that could be fraudulent or do not subscribe to AML requirements.
Let’s take the process of verification of documents presented under a Letter of Credit (LC) from foreign banks. When done manually, not only is there a longer transaction turnaround time, but the process suffers from high error rates, lack of standardization, and operational risk due to staff turnover. The AI solution extracts relevant information from trade documents and matches terms and conditions in the LC with actual data in the trade document. Based on the findings, discrepancy advise is triggered to the exporter’s bank if relevant. This can potentially result in 70% savings in average handling time.
Execution
Once a decision is made, execution, or the actual implementation of the decision is the last and most critical step. Digital workers, for instance, can be leveraged to execute steps such as reading inputs and decisions from various sources and capturing the request in the system, creating payments, reconciliation etc.
By reimagining banking processes through AI and Automation, banks can benefit from a “human++” model, where employees’ can deliver greater value. AI can free up employees’ time drastically to perform more cognitive tasks and enable them to deliver greater value through cognitive tasks.
