Computer Merchants undertook a series of innovation projects which spurred the creation of two new tools: CM View (for automated system maintenance), and CM Care (for customer asset management).
Both applications are hosted on IBM Power Systems servers running the IBM i operating system, as Norm Jefferies explains: “IBM i and IBM Power Systems technologies have been at the heart of our infrastructure for a very long time—since the 1970s, in fact! During this time, we have become very familiar with the systems, and we know how to operate them well. By choosing IBM i to develop our new applications, we knew that we were going to benefit from a reliable, stable, and established system, which would enable us to ensure the continuity of our business during these times of change.
“Additionally, we chose to deploy IBM systems because we knew that they would enable us to run our traditional business model alongside cutting-edge technologies such as analytics, automation, and machine learning.”
CM View is a tool that Computer Merchants initially created to help it monitor its own internal systems, which the company then developed further to help support its clients’ infrastructures. Featuring free mobile and tablet capabilities, the application provides a consolidated, real-time view of system performance, so that Computer Merchants can proactively offer assistance to its clients whenever an issue arises.
Norm Jefferies notes: “CM View started out as an internal tool for Computer Merchants, which we offered to a handful of our clients as a free trial—they loved it, so we developed it further, creating an easy to use graphical user interface [GUI] that gives our clients a view into the health of their systems. Now, our fully developed CM View tool totally automates machine maintenance for our clients, by detecting degradation in service levels, flagging up potential root causes, and keeping our clients and engineers notified.
Computer Merchants is enhancing CM View with IBM Artificial Intelligence capability. It is also leveraging open-source machine-learning algorithms from the Python Scikit-learn library which the company runs on IBM Power to train a Decision Tree Classifier (DTC) with past outcomes. CM View will help Computer Merchants anticipate the behaviour of maintenance tickets in future with better than 90 percent accuracy. This figure is set to rise as the system analyses more data.
Norm Jefferies adds: “The machine-learning technology underpinning this new version of CM View means that we can see which tickets are most likely to need assistance from an engineer, and which tickets can be solved automatically, which helps us to optimize our use of human resources.
“Building on the success of CM View, we developed CM Care—an asset management tool also running on IBM i. This free tool provides our clients with a view of the status of the support or maintenance contract they have with us, so they know exactly when their contract will expire and when to take necessary action. In the same portal, we also display exactly when a ticket has been created and when action has been taken by one of our engineers. This way our clients not only have a transparent view into the health of their systems, but also into the action we have taken to support their infrastructure.”
Eager to find new ways to improve customer service, Computer Merchants launched a project to integrate customer emails from its customer relationship management (CRM) system with an application program interface (API) on IBM Watson®.
“First, we separate customer emails from non-customer related mail, such as emails from our suppliers and vendors,” says Norm Jefferies. “Then, we send those customer emails up to Watson Tone Analyzer in the cloud to determine how satisfied or unsatisfied a customer is, distributing these results across a scale of positive-negative.”
“The negative customer interactions are flagged up, so we can look into the issue, identify trends and transform processes. Using this tool, we found that most customer dissatisfaction originated from the fact that we simply didn’t return or answer customer calls. Unlocking this type of insight is crucial if we want to deliver excellent service and assistance to our clients at all times, and minimise the risk of having dissatisfied customers in the future.”