Machine Learning: More Than a Buzzword

Machine learning has been defined by Stanford University as “the science of getting computers to act without being explicitly programmed.” While the past few years have seen a rise in demand for conferences, venture investments, and business-related inquiries, many technology executives are still exploring the wide-ranging applications of machine learning in business issues.

Image source: pixabay.com

Sales data analysis is one of the many contributions of machine learning to business. It has benefited from the growth in sales-centered data because of the increase in digital interaction. Through this, sales teams have the ability to tap into metrics from social media platforms, website visits, and A/B testing without having to experience systems bogging down on them due to the massive volume of data they have to sift through. Notably, machine learning speeds up the process of revealing the most useful information. 

The fraud detection process has been significantly improved as some payment services employ machine learning tools to search for fraudulent transactions such as money laundering. Machine learning processes segregate these from lawful transactions. It examines specific features in a data set and then establishes models which provide the foundation for reviewing and classifying each transaction. This assists in preventing fraud even before the completion of the transaction. 

While many believe that AI takes the job of compiling reports and making decisions from employees, that isn’t necessarily the case. Data-entry employees can still use their experience to deliver value to firms that no longer need as many hours of manual data scanning. 

Image source: pixabay.com

Edmund O’ Brien is a certified public accountant and serves as the CEO of the Americas at Arvato Bertelsmann. His most recent responsibility has seen him in several executive positions at Microsoft, including Vice President World Wide Operator Channel. For similar updates from Edmund O’ Brien, subscribe to this blog.

The Upsides and Downsides of Serverless Computing

Back in 2014, serverless computing gained huge popularity when Amazon first released the AWS Lambda. Several vendors, an example of which is Microsoft’s Azure Functions, followed suit by developing their own serverless computing solutions that sought to address the varying needs of consumers or organizations.

Image source: cuelogic.com

A serverless computing solution offers the following advantages:

• Decreased costs: There would be lessened operational and development costs because the organization can opt to use only the resources that they need.
• More manageable cost to scale: Developers do not have to implement code to scale, in addition to eliminating the requirement of an upgrade of servers should the organization find the need to increase the used cloud computing services.
• Reduced complexity: Serverless computing does not need as many interfaces as possible, including virtual machines, servers, operating systems, and more.

However, there are also drawbacks to using serverless computing. And while upsides definitely outweigh the downsides, the latter is still worth noting:

• Possible performance issues: Because the allocated compute resources are dependent on the applications’ requirements, there might be high latency, which can affect performance. If high performance is a primary requirement, allocated virtual servers are a better alternative than serverless computing.
• Tricky monitoring and debugging: In serverless computing, multiple server resources are used. Due to this, monitoring and debugging activities can be difficult to do.

Image source: trendsbuzzer.com

Eddie O’Brien has extensive experience in global sales and services, technology, international business strategy, and cross-functional leadership. He currently heads Arvato Bertelsmann’s Global Cloud Strategy. He had previously worked for Microsoft for 15 years. Follow this Facebook page for more discussions on technology.