Development policy

Various RPA needs as received from clients

Various requests for RPA related solutions are sent to Net Smile. For example,

  • Is it possible to automate e-mail interactions with customers for a real estate company?

  • Would it be possible to suggest an optimal diet based on the employees’ current health conditions, in order for company employees to better maintain their health?

  • Can we automate the process of determining the strength of residential land soil for a ground analytics company?

  • Would it be possible to create an investment AI that automatically buys and sells shares for investment and other types of companies?

  • Is it possible to automatically translate the sentences written by takers of English exams into digital text?

  • When financial institutions are considering lending to someone, can we automatically determine the risk of loss based on the customer' s attributes?

... and so on; a great variety of inquiries.

Does AI (Artificial Intelligence) require 100% accuracy?

Clients with goals like those listed above obviously hope to achieve 100% accuracy. However, it is difficult for AI (machine learning) to achieve 100% accuracy in actual complex tasks.

For simple tasks, 100% accuracy is possible with a rule-based system, but the AI is expected to learn complex business processes in which even human beings make mistakes.

With Net Smile, rather than pursuing 100% accuracy from the beginning, we introduce human-like AI that incorporates mistakes to the same extent that would arise artificially. “Training” the AI then improves its accuracy and results in the reduction of total costs.

Focus on implementation speed and “training”

The more complicated the work, the more likely humans are to make mistakes. However, it is possible to guarantee the accuracy of the final output by creating a plan in which humans and the AI check each other.

When AI with this type of system is used, the goal of cost reduction by RPA can be achieved.

By introducing an AI RPA system that includes a few mistakes and set it to work as early as possible, we can ensure that there is time to be used for "training." The reason for focusing on implementation speed and training is the reduction of costs through RPA.