What are the barriers that hinder the use of DX/AI?

According to the "DX Report 2 Interim Summary" released by the Ministry of Economy, Trade and Industry in December 2020, 95% of companies are either not working on DX at all or have just started working on it, and there is a company-wide sense of crisis. We have not reached the stage of sharing or changing consciousness. And we analyze that there is a big difference in the progress of DX between leading companies and average companies.

Most companies want to promote DX, but it is only sporadically implemented by each department or stops at the information gathering stage. We are aiming to establish a system for business transformation and become a digital company that can respond quickly to changes in the external environment, but it is not going well.

There are three major barriers to the shift to a digital company: formulating a DX strategy, developing a DX promotion system, and securing DX human resources. Unless we overcome these barriers, we will not be able to reach the original purpose of DX, which is “business transformation utilizing digital technology.”

On the other hand, what kind of efforts are being made by companies that have successfully overcome these barriers to digital transformation? At a major American retail company, robots equipped with cameras and cameras installed on the ceiling take pictures of product shelves, and by analyzing the accumulated image data with AI, it is possible to determine in real time which products are depleted, at what timing, and by how much. is calculated to

As a result, data-driven measures can be taken. For example, an app for customers sends a map of the store and the status of out-of-stock items on the product shelves. This will prevent the situation where you actually go to the sales floor and the product is not there. From the customer's point of view, it will be a highly convenient store that "because the product inventory is visualized, you can always know that there is the product you want when you come to the store." In other words, data and AI are changing the customer experience itself. In addition, product shelf data is linked with warehouse inventory data, which is also used to optimize the supply chain.

The point is that data is the source of competitiveness, creating new value by using AI, and creating a loop that circulates them. I believe that DX is not an effort to digitize with pinpoint accuracy, but rather a loop that creates new value while linking the entire business in this way. How quickly we can create this loop will make a difference in our competitiveness. Therefore, the important things are "DX strategy", "DX promotion system" and "DX human resources". The most critical issue among these three is the lack of human resources.

Three points to solve the problem of shortage of human resources

Efforts are being made to create new positions such as CIO and CDO, appoint them from outside, and think about strategies centered on those people. Human resource DX skills and literacy are overwhelmingly lacking. What skills should the workforce have in the field?

Roughly broken down, skills in areas such as business power, data engineering, system engineering, and data science are required. Of these, when it comes to utilizing big data and AI, you will need data science and data engineering capabilities. It is said that there is a particular shortage of human resources who can handle these areas, and how to supplement them is directly linked to the success of DX.

There are three ways to compensate for this shortage of human resources: "Recruit new personnel," "Upskill internal personnel," and "Use external resources."

New recruitment is difficult to procure and there is also a problem in maintaining human resources. It takes time to improve the skills of internal human resources, so although it is a measure that should be taken in the medium to long term, it is not possible to reach immediate results. Using only external resources has an immediate effect, but it is necessary to select an appropriate vendor, and it is necessary to prevent the situation where the vendor is left unattended and cannot run on its own. Rather than relying on one or the other, it is necessary to consider all of them in a balanced manner.

However, attracting data scientists with the right skill set is not an easy task. Therefore, Macnica proposes finding a "co-creation partner" and working together to advance DX. DX is the use of digital technology to create new value that is not an extension of the past. To achieve this, we do not need to be self-sufficient, such as ``complete everything in-house,'' but we believe that we should achieve this through co-creation activities, making good use of knowledgeable business partners.

The knowledge, data, customers, and ``mastery of digital'' that a business company has cultivated over the years will definitely be needed. In addition, we are required to utilize the technology, technology, and knowledge of our IT vendor co-creation partners, and to think from the perspective of what can only be done because of digital technology. We believe that digital transformation is what is created through medium- to long-term co-creation activities by combining these factors.

Utilizing the data scientist community is effective in solving problems using AI

Next, we will introduce some points when incorporating AI, which is essential for DX, into your business.

Businesses differ from company to company, and the data they generate also differs. Depending on the differences in your business, the appropriate co-creation partner will also vary. Additionally, each company has different DX strategies, DX promotion systems, human resources, and KPIs that serve as evaluation standards. Therefore, there is no AI that can produce high accuracy and business impact under any environment or constraint conditions. You need to think about and create the best AI usage and model for your company. There are, of course, differences on the vendor side as well. Vendors have different human resources and areas of expertise, so it is necessary to select a co-creation partner while assessing your company's situation, issues, and strategies. If your company is in the phase of developing an AI model, one of the best practices for AI development is to utilize a "data scientist community."

By utilizing the data scientist community, we can select the model with the highest evaluation for the problem, dataset, and evaluation index through a competition. If you have an AI model development proposal linked to a business problem, submit the problem and dataset to the community. Then, data scientists from all over the world can compete and simultaneously develop AI models that lead to the optimal results for the problem. By comparing different approaches to feature design and the knowledge of data scientists around the world, you can choose a scientific method.

AI models are developed using a so-called crowdsourcing approach, which is said to be one of the most effective methods. There are 3 reasons.

The first is that the data scientist community has a diverse range of people, so we can gain the wisdom of 100 or 1000 people.

The second reason is that we can generate innovative developments through open innovation. It is more likely that ideas for better solutions will be generated if you gather talented people, not just your own people, to consider problem-solving.

The third reason is that many different approaches can be compared in a short period of time while avoiding data scientist bias. For these reasons, we believe that AI model development utilizing the data scientist community is excellent.

At Macnica, we support companies' digital transformation from various perspectives, centered on CrowdANALYTIX's data scientist community. Please contact us for details.

▼ Business problem-solving AI service that utilizes the data science resources of 25,000 people worldwide
Click here for details.Fully customized AI service CrowdANALYTIX