Company Introduction

ShareWith Co., Ltd. is an e-learning company headquartered in Osaka that develops and provides the LMS "WisdomBase."

The company's main customers are a wide range of organizations, including universities, corporations, and other organizations, and it handles a wide range of tasks, from selling e-learning materials to certification exams and in-house testing. Its greatest feature is its high level of customizability, allowing for flexible system configuration according to user needs. It places emphasis on user-friendliness in its UI/UX, and is highly rated for its ease of implementation, even for companies that find it difficult to develop their own systems in-house. In the 17 years since its founding, the company has been combining education and digital technology to continue to innovate the way learning platforms are provided.

Issues before introducing generative AI

Although generative AI was already being used in some operations, the reality was that it was not being used to its full potential.

In field sales, AI was used to create follow-up emails after sales negotiations and to prepare meeting minutes, but it required repeated "bounce-backs" to ensure the system understood the intentions, which limited the efficiency of the system. Furthermore, when creating blogs, the team struggled with low output accuracy due to their own prompts. It took more than three hours of struggling in front of the computer just to produce satisfactory output, and then further work was required to confirm the content and refine the text, so it took a huge amount of time to complete a single article.

This personalization of skills and inefficiency at the field level, where people have the tools but don't know how to use them, was a challenge for the entire organization.

Effects after generative AI introduction training

We have evolved from "trial and error" to "purposeful use." Learning the basics of how it works has enabled us to design precise prompts, and has cut the time it takes to create minutes and emails by more than half.

Even when creating blogs, AI can quickly generate high-quality "rough drafts," significantly reducing the amount of work required to publish them. Previously, it took more than three hours just to get a response, but now, including final confirmation, it only takes about three hours in total. This has enabled humans to focus on essential tasks such as verifying and revising information.

Even more than just numerical results, there has been a major change in the mindset of employees. By learning about the depth of AI in the training, employees have established the habit of proactively keeping up with the latest information. By understanding the importance of prompt design, they are now proactively thinking about how they can utilize AI in tasks where it has not yet been introduced, and have evolved into an attitude of wanting to expand the scope of application.

Future outlook

We believe that it is important to standardize and fit the knowledge gained through training into the company's business, rather than leaving it as an individual skill.

First, the members who learned as pioneers this time will become leaders in promoting the use of AI within the company and strengthen knowledge sharing at a practical level. Specifically, they plan to roll out advanced use cases, such as research using Gemini and automating slide creation, throughout the organization. They also plan to attempt to restructure business processes, including in the development field, by using Dify to integrate systems with Slack and HubSpot, with the aim of drastically reducing the amount of time spent communicating within the company.

By leveraging the deep understanding gained from the training regarding the importance of security risks and information management, we will promote the safe and effective use of AI, thereby further improving productivity and accelerating the provision of value to customers as a company that provides an educational platform.