e-Learning Blog

Trends by Project Type (Compliance / Skill Transfer / Knowledge Acquisition)

We categorize the projects our company has handled since 2025 into three types and share the AI utilization ratio by process and the resulting image of changes in production man-hours.

① Product and Service Training, Skill Transfer: Training centered on company-specific information

Education that deals with company-specific knowledge and know-how is the most challenging area for AI utilization.

How to share internal knowledge and technology cannot be sufficiently generated using only the general information that AI has learned. It is realistic to proceed with many processes manually while utilizing AI in some auxiliary processes.

In such cases, documented information is often limited or not sufficiently organized in the first place. Many areas rely heavily on the knowledge and experience of instructors and on-site personnel, so it is necessary to first extract information through interviews and then reorganize the dispersed materials into educational content.

Furthermore, it is necessary to take into account company-specific terminology, product characteristics, and on-site precautions unique to the field. Additionally, it is essential to identify who the target audience is and determine which types of expressions (such as diagrams, illustrations, animations, interactions, etc.) are easiest for them to understand, designing through repeated practice and review.

In this area, AI is effective for assisting with material creation and information organization, but human involvement is indispensable in the core processes of educational design.

Also, because this area relies heavily on manual work, production costs tend to rise easily.

On the other hand, when there are many types of products or a large number of skills to be passed down, it can be more effective to establish a system within the company to accumulate and share knowledge rather than creating content each time.

It is necessary to consider multiple options, including not only the utilization of AI but also the introduction of knowledge-sharing systems, the development of shared templates, and the design of operational rules.

② Store Customer Service, DX Training: Training that mixes general knowledge and company-specific circumstances

In this area, while there is general knowledge in the foundational parts, the business workflows and system operations actually used on-site contain a lot of company-specific information.

Therefore, although AI is somewhat easier to utilize than in ①, the final quality still depends on human judgment.

In such cases, the implemented systems and operational rules themselves are highly unique, making it difficult for AI alone to properly organize how to use them.

In our company, especially when dealing with operational procedures, technical writers determine where learners are likely to get confused and translate that into manuals and video expressions.

Also, regarding business workflows, even if the differences between the current paper-based operations and the post-system implementation operations have been organized, a lot of ingenuity is required in how to communicate this and how to ensure it permeates the workplace. Expression and implementation design such as animating the content and combining multiple educational measures are also important.

AI is effective at generating draft plans for structure and expression ideas.

On the other hand, the task of deciding how to actually use these and shaping them into a form that is understandable to learners still primarily relies on human roles.

③ Hierarchical Training, Information Security, Business Skills, Harassment, DE&I: Training centered on general knowledge

*However, when dealing with company-specific cases or decision criteria, it approaches ② or ①.

This area is the easiest type of education to utilize AI among the three.

Since the foundational knowledge and general explanatory frameworks are somewhat common, by conveying specific objectives and target audience profiles to AI, it can create drafts for many processes.

However, while AI excels at quickly presenting generalized safe expressions and standard structures, it cannot automatically produce content that deeply fits the learner's situation or designs that lead to behavioral change.

Even in education that conveys generic information, the educational content and expression methods vary greatly depending on the target's situation, the reason for conducting the education, and the goals for what should happen after completion.

It is true that many processes can be entrusted to AI, but verifying the accuracy of the content, adjusting expressions to suit the target audience, and ensuring consistency with learning objectives need to be handled by humans.

Also, when it is sufficient to add just a little of your company's unique supplementary information to general information, not only creating new content using AI but also adding your own information before and after existing sales content is an effective method.

With this method, it becomes easier to secure a certain level of educational effectiveness while keeping costs down.