Webinar: Why Can't We 'Master' Generative AI?

## Accelerated Adoption of Generative AI, Emerging Barriers to Practical Implementation

As the use of generative AI expands, while some PoCs or departments have seen success, many companies are struggling to scale these efforts company-wide, often finding themselves reverting to old habits. Even when implemented to fit departmental tasks, results are often not reproducible, success patterns are not captured, and initiatives remain fragmented.

The issue lies in the fact that knowledge is not captured in a "lasting form." Generative AI usage requires not only providing tools but also continuously maintaining and updating essential operational knowledge such as prompts, procedures, criteria, prohibited items, and data handling. It is crucial to look beyond simply blaming the lack of end-user engagement and instead re-evaluate the infrastructure for internal knowledge retention.

## Lack of Rules Hinders Company-Wide Expansion

In reality, initiatives often proceed without established guidelines or rules, leading to fragmented operations across departments. Consequently, concerns regarding information management and security limit the scope of use, stalling discussions on company-wide expansion.

Furthermore, when knowledge remains confined within individuals or teams, it is lost upon personnel movement, retirement, or project termination, leading to repetitive trial and error. Moreover, proceeding with ambiguous requirements means it is unclear "who is using it for what business purpose," making evaluation and improvement impossible. To achieve sustained results with generative AI, the challenge is to pair requirement definition with operational design to create a mechanism that leaves behind "usable knowledge."

## Hands-on Support from Guideline Formulation to Adoption

This seminar will explain how to build "usable knowledge" from the perspective of requirement definition and operational design to expand generative AI usage company-wide and continuously increase success cases.

Specifically, we will introduce concepts for maintaining continuous field usage and constant knowledge updates, while covering key points in establishing guidelines (e.g., clarifying usage goals, data handling, and setting operational rules). Additionally, to ensure efforts do not end at setup, our engineers will share their practical experiences on how to provide ongoing support and drive improvement cycles until the system is fully embedded.

We invite those who wish to scale generative AI initiatives from fragmented points to a company-wide practice, or those looking to resolve issues where internal knowledge is not retained and cannot be shared, to participate.

■ Organizer
TechnoPro Holdings, Inc.

■ Cooperation
MajiSems, Inc.