Mediaboss Marketing

Before innovating any product or solution, it is essential to understand the pain points or challenges of the audience. Conversations with people in the audience help towards that goal, and the Samsung SDS team actively seeks them out to learn from them. Often, the challenge becomes clearer only when one understands the practical daily life of the users.

Patrick Bangert is the Vice President of Artificial Intelligence at Samsung SDS, America. He mentions that once, he was asked to construct a damage forecasting model for a certain kind of pump to lower maintenance costs. At that time, he lacked experience and was asked to do a three-day internship with the maintenance crew to find out. They took him along, and he performed the work with them; all the while experiencing what they go through. In the end, he had a clear picture of where the problem lay and then could solve it much more easily.

Patrick started his career as an Assistant Professor of mathematics at a new university that had only existed one year before he started working there. There was no curriculum set up for the courses that he was supposed to teach. So he started by putting together a coherent end-to-end program and then explaining it to a large audience of undergraduates. It was a difficult lesson that forced him to explain things in simple terms. The dual skill of making a complex subject accessible in simple terms and breaking it down into digestible pieces are crucial to leadership and inevitable sales and marketing function as a leader.

His challenges did not end there. Patrick explains how at the beginning of his journey, he did not know where to begin and lacked a clear vision of where he wanted to end up.

With hindsight, I recognize that having a vision of where you want to be in your own life at an older age is a boon because it acts as a guiding principle in decision-making.

Patrick recalls several people guiding him and offering him suggestions to do this and not do that. He followed the advice most of the time because he respected them but did not necessarily understand why, except for short-term tactical reasons. Naturally, this led him to make a few costly mistakes, the worst being choosing the wrong business partner in founding his startup company. He did not know how to perform proper due diligence and did not do it. His partner was unable to fulfill his responsibilities, creating a profoundly serious problem for the business. Subsequently, he was forced to change his role and fulfill both leadership functions at the same time. For him, this was an existential challenge, but he learned a great deal.

Currently, under his leadership, Samsung SDS is leveraging disruptive technologies to give the company a competitive edge in the market. These technologies make it easier and faster to innovate and enable several innovations themselves. Recently, Patrick has been working on using AI to diagnose diseases based on medical images. He says that creating these models is very effortful, particularly annotating the images to generate the dataset; on which to train the models. Using AI to help the annotators, the company could lower the labor by over 80%, making it significantly faster and cheaper to make models. The company is calling this technology AI2 since it is using AI to create more AI.

If we categorize human history by its most influential tool into stone, copper, bronze, and iron ages, then the current time has sometimes been called the information age. While the information itself is an asset, it is only really the conclusion of an analysis that adds value. So, currently, we live in the analytics age with a ubiquitous impact on further innovation, he said.

According to Patrick, there have been many changes to human society in recent centuries. The current AI-driven transition is the so-called fourth industrial revolution. There is some fear attached to the inherent change, as in any transition can make some jobs less needful, but other new job functions usually take their place. There is less and less need, at a grand scale, for manual labor. In the industry overall, the transition towards using AI is still in its early days. AI is mature only in the retail industry and in some isolated applications elsewhere. There is a long way to go to adopt the analytics we have. Those who have developed the analytics must work on making them better. The primary directions are AI ethics, explainability, and the challenge of small data.

As a large IT services company, Samsung SDS is the chief player in this transition and the importance of analytics grows both internally and externally. The company will keep developing new tools and models to serve the needs of the global business community. At present, it is actively focusing on the challenge of small data: the fact that often the dataset has a limited size and cannot be increased readily, so AI must make do. A host of new algorithms are needed to make good quality models once the standard quality improvement mechanismget more databecomes impossible.

As a Ph.D. mathematician, Patrick started to work in academia as a Professor in Germany and proceeded to found a startup in the field of machine learning as applied to the oil and gas and chemicals sector. During this time, he led projects all over the world to do predictive maintenance and process optimization for almost all the major oil companies. Having grown that company to maturity, he was hired as the VP for AI at Samsung SDS, the IT company of Samsung Group.

At Samsung SDS, he leads three teams. AI Engineering makes software tools for data scientists, primarily in speeding up the process through distributed training and AutoML. AI Sciences does the model building within Samsung Group and for outside customers. The third team does the sales and marketing for the other two teams. Besides the companys regular work in natural language processing, computer vision, and time-series modeling, it has recently shifted its focus to medical imaging.

Read more here:
Patrick Bangert- Leading the Artificial Intelligence Industry with Disruptive Innovation - Analytics Insight