Demo
On the one hand, the learning machine is able to predict the correct category for the operators to move to the automatic workflow of the next process while determining the behaviour patterns in the process they are doing.
Because financial institutions want to implement the best solution, they learn to reduce their operational costs and human operational errors to the minimum level with the aid of learning machine technology. Organizations such as call centers, customer service management and branches with intensive transaction volumes in the banks have a high risk of dramatic work loss and operational error. What is expected from a customer representative that meets hundreds of calls a day at a call center that millions of customers are looking for is to direct the request by triggering the previously planned related workflow if it is not able to provide the right solution by categorizing the claim of the customer accurately and correctly at every call.
According to a study at Cornell University (☛), an ordinary person makes decisions only about eating approximately 226.7 times per day. Some internet sources claim that people make 35,000 decisions a day. Though it is assumed that a significant part of this is unconscious, it is still said that a person makes a decision about 2,000 a day.
How many operational decisions do you make for a professional white-collar job? More probably …
The question that should be asked is: do the number of mistakes in these decisions increase as the number of decisions increases?
Banks have to minimize their error risks and constantly monitor their operational efficiency. For this reason, they use a simple but very effective helper function that the learning machine provides to them, especially when they classify issues they receive from customers. When the customer representative starts to submit the customer’s issue in the customer service management application by pressing the keys of the keypad, the suggestion of which type of transaction is relevant is automatically determined with 90% accuracy. The trick here is that a cognitive pattern is detected between the characters in the text of the learning machine and the behavior of the customer representative is related to this pattern through sensors.
We all make mistakes in some of the thousands of decisions we make in the day. But we do not really care because most of our mistakes are related to ourselves. On the other hand, the operators who decide on behalf of others, customer service workers, call center operators? They need to use the blessings of technology to minimize their mistakes.
Contrary to popular belief, artificial intelligence or learning machine technology will never be able to take over the call centers or operational units. However, it will more directly contribute to the increase in labor productivity by reducing the risk of operational human error that could arise from these large structures.
Gurkan Platin,
July 2018
