by Terence Leung

An article by the Forbes Technology Council (“Machine Learning And Analytics: What’s Your First Step?”) takes a look at recommended steps to start exploring Machine Learning. The article mentions that, as an early step, there is the need to start with connectivity and data collection. It also highlights the importance to have clear goals.

In helping industrial customers on improving business performance and this machine learning (#ML) journey, I find that machine learning is not widely adopted yet on the business management level. Whereas on the shop floor, warehouse or customer support, artificial intelligence and machine learning are being deployed for very specific purposes, such as optimizing machine maintenance schedule, further automating pick-and-pack, and conveying contextual information to customer service representatives.

With this kind of excitement multiplying (especially this year!) and with initial success on the above use-cases, executives are now looking at applying machine learning at a strategic level.

Connectivity and Technology Considerations

For business management, such as management of profitability and cash flow, many organizations leverage  business intelligence tools as well as reporting functionality from their enterprise systems. To analyze a new business or situation, they may have a problem of lack of data. To overcome this hurdle, they have to ask their IT departments to configure appropriate reports and finance personnel to set up financial analysis. The process can take weeks or even months. By then, situations may have changed.

For machine learning projects to be successful and continue to be fruitful, data infrastructure is a major consideration. How can we get the data and timely updates from various systems to solve a particular problem? To circumvent such challenges, organizations are leveraging big data technologies to load in data from an extended number of systems.

Still, organizing the analysis will need the help of trained analytic resources (such as data scientists) and/or outside consultants. Short-term shortage of such resources is an immediate concern. More importantly, upfront investment and business cases are essential considerations, sometimes not just at a departmental or regional level, but corporate level.

In relating to the Forbes article, having clear goals, such as performance objectives and process improvement benefits, is essential. I will go even farther in advising that: as sophisticated business executives, the bottom-line question to ask upfront is, “ how much time, effort and investment are needed to achieve such benefits”?

Writing checks for an experiment is so 90s……With the proliferation of cloud solutions for specific purposes, the build-versus-buy equation has to be considered. I find that the benefit to shop around is really just for the best value; in doing the necessary research, executives will learn a lot from the success stories and the use cases that tend to be providing benefits with the current state of the art technologies and/or solutions. Regardless, by getting your business and technology teams involved in the research and decision-making process early on, an acceleration effect could be achieved, as everyone will understand what machine learning will (and will not) accomplish in terms of business value.

On the connectivity and technology front, we at Greenlight Technologies are working toward a more flexible paradigm, leveraging our off-the-shell deep integration with 100+ enterprise systems and advanced analytics platform, we enable executives to use predictive analytics and machine learning without the headaches of data acquisition or the need for a lot of resources. Our roots started from continuous monitoring of risks and their financial impact. We provide our customers with 360° visibility for virtually all the systems they have. Therefore, the set-up time and effort are drastically reduced.

Look for part II soon

A note about the author: Terence is designing analytics solutions to enable Greenlight Technologies to serve the increasing needs of customers on decision-making related to processes, operations and transformations. He was previously at Deloitte Consulting’s Finance, Operations and Strategy practice and at solution providers including i2 Technologies that optimize company performance and processes. He really enjoys interacting with industry practitioners on topics such as business models, value, technology and especially analytics.