Paving The WayTowards Innovation

CASE STUDY

Segment

Industry
Manufacturing

Challenge
Smart Search for Emails

Practice Alignment
Machine Learning

Decades of Old Data Suddenly Becomes Available

Cy.Pag. is a leading international manufacturer of pneumatic cylinders for industrial automation. In addition to this, they also manufacture hydraulic and electric cylinders. The company was established in 1995 in Valtellina, Italy and now stretches over an area of 43,000 square feet. Cy.Pag. has been in business with thousands of customers over the years and, like most of their competitors, relay import product scenarios via email communication with their customers and distributors.  These product scenarios are often cyclical and re appear with their customers or new customers that come along.  

Although recurring business was good, the employees of Cy.Pag. found themselves searching through hundreds of emails to find similar situations and solutions they have used with past customers or sales opportunities.  The important information they were looking for was buried in email communication that was relayed months or even years ago.  In order to bypass this, they sometimes would turn to the only one who knew this by heart, their lead engineer.This caused concern for company CEO, Luca Pagni.  Most of the solution information was inside the mind of their lead engineer (or lost in email) and he wondered what they would do if he left the company. Specific solution scenario information was only in his head, or buried in thousands of emails.The employees tried using the built-in search that is supplied in all email inboxes, but they found this ineffective. 

According to the employees of Cy.Pag., the built-in email search was ineffective because it utilized key-word search.  Oftentimes, the information details were not in mind; this is why they found themselves asking lead engineers specifics about a past solution.  

They then began to hear of a machine learning concept called “natural language understanding” (NLU).  After consulting with some cloud exports, he was offered a knowledgebase built on email archives.  The product allows users to naturally query their inbox (or company designated set of emails) to find the information that they need.  The solution empowered the employees to query their email with just “the gist” of what they’re searching for and receive results that were spectacular.  Also, they were now able to ask ebase.ai natural questions, just like they did with their tenured engineer.

Outcome

The product allowed for multiple users and an administration feature that allowed a company executive to add new users and remove old ones.  This web application is secured through Amazon’s Cognito service that provides authentication and authorization.  According to Luca Pagni, ebase.ai provided peace of mind.  He knew now that they can treat their email inbox as a trusted source of information that they can ask, as opposed to their lead engineer (who can now focus just on his work).  Also, the tool increased agility in the company and empowered their sales and partner relation employees to turn around business more than twice as fast as their normal speed.  Cypag is now growing faster than ever and has even released new company subsidiaries such as SIS (Safety Improvement Systems) with smart forklifts.  The ebase.ai team is happy to be a part of that success made possible by ebase.ai.

Technologies Used

Amazon SageMaker

AWS Lambda

Amazon S3

Amazon API Gateway