Digirise Inc. (Headquarters: Minato-ku, Tokyo; CEO: Masahiro Chaen) has implemented a large-scale update to its enterprise-oriented ChatGPT "AI Works."

◾️ Overview of the major AI Works update: A complete list of new features and enhancements that will accelerate productivity improvements in companies 

In addition to building a cutting-edge infrastructure by adopting Google Cloud Platform, this update also brings about enhancements to further accelerate corporate digital transformation, such as the addition of new functions that promote the use of AI throughout the organization.

The main updates this time are as follows:

• Adopting the latest Google Cloud Platform to build a more advanced service infrastructure for businesses
• Create templates to facilitate knowledge sharing across your organization
• A new UI design has been introduced that allows for intuitive operation, including the adoption of a card-type list display.
• The addition of file attachment functionality to the GPT-4o mini model enables more flexible document analysis 
• Flexible character limiting based on business needs and introduction of a new, highly transparent billing system

One notable new feature is the addition of an internal template sharing function, which makes it easier to accumulate and deploy effective AI utilization know-how within a company. In addition, the adoption of the latest cloud infrastructure ensures stable performance even when used by large organizations.

Through this update, Digirise aims to encourage more companies to utilize the latest AI technology and achieve further business efficiency.

◾️Evolution of infrastructure through migration to Google Cloud Platform: Achieving a highly secure security posture

Digirise has migrated the service infrastructure of AI Works to Google Cloud Platform (GCP) in order to realize a more advanced AI platform for businesses. By adopting GCP, we have built a comprehensive system foundation, including linking with OpenAI API, integrating File API and Assistant API, authentication system with Firebase Auth, and management of chat information with Fire Store.

Security during transmission and storage is particularly important when protecting corporate data. In GCP, data is automatically encrypted when written to disk and protected by splitting it into multiple chunks. This encryption key is strictly managed by Google's key management system, and TLS is also applied when transmitting data to ensure security during communication routes.

Features of GCP-based infrastructure

• Use of advanced communication protocol via HTTPS 
• Using the Firebase Auth authentication system 
• Optimized access management for each organization
• Utilizing GCP's automated encryption system 
• Use of encryption protocols during transmission

In addition, high reliability is ensured for data exchange via API. It is guaranteed that data sent via the API provided by OpenAI will not be used for learning, and the same security policy is applied to file transmission. This ensures the protection of important information assets of companies.

The migration from Azure to GCP is carried out in three steps:

1. Once migration is complete, complete migration to the new system will be implemented.
2. Maintain monitoring by Digirise employees for the first six months after the transition
3. After six months, all migration-related data will be completely erased.

This migration process ensures the safe migration of data while ensuring the continuity of services. A six-month monitoring period is set up after the migration is complete to thoroughly check the stable operation of the system and the integrity of the data. Furthermore, after this period has passed, all data related to the migration is completely erased to ensure thorough information management.

◾️ Promoting the use of AI throughout the organization: Details of the new "Organizational Template Sharing" feature 

To promote the use of AI more efficiently within a company, we have introduced a function that allows templates to be created and shared within an organization. In addition to the traditional individual-based template creation, we have created an environment where knowledge can be shared across the entire organization, which will improve the work efficiency of the entire team.

The new template function uses an intuitive card-type list display, allowing members of an organization to easily refer to and use templates organized by business category.

Main features of the template sharing feature:

• Improved visibility by displaying card types at a glance
• Efficient template sharing among organization members 
• Ability to accumulate best practices within an organization 

In particular, in terms of practical use in the enterprise, we envision the following use cases:

• Share standard prompts for routine tasks 
• Deployment of department-specific, specialized AI utilization know-how 
• Efficient knowledge transfer to new members 
• Accumulation of best practices for using AI across the organization

These functions enable each organization to efficiently share and deploy know-how on how to utilize AI, accelerating the promotion of digital transformation across the entire company.

◾️New UI design for improved convenience and operability: Creating a more intuitive operating environment

We have completely revamped the UI to make AI Works a more user-friendly tool. From improvements to the sidebar to changes to the way templates are displayed, we have designed it with user convenience in mind.

One of the biggest changes is the improved sidebar to improve access to the template function. The buttons to return to the home screen have been reorganized, and the template button has been newly placed, making it easier to access the functions you need.

Key UI/UX improvements:

• Improved operability by displaying template usage screen on the left side
• Improved visibility by introducing a card-type list display
• Improved accessibility by reorganizing sidebar functions
• Efficient management and sharing of templates
• Added file attachment feature to GPT-4o mini model

This new UI also includes improvements to the file attachment function. File attachments are now possible on the GPT-4o mini model, enabling more flexible document analysis. Users can attach files as needed to facilitate smoother communication with AI.

◾️Flexible character limit and new billing system: Develop a usage environment that meets business needs

In order to accommodate the diverse usage needs of companies, we have improved the way AI Works handles character limits and introduced a new billing system.

We have implemented more flexible character limits. For example, even if you are using 2,999 characters and the limit is 3,000, you are not limited to the remaining 1 character, but can enter the required number of characters. This prevents interruptions to business and allows for a smoother usage environment.

◾️About Digirise Inc.<https://digirise.ai/＞

Digirise Inc. started its business in 2023 with the vision of "bringing the convenience and excitement of AI to everyone." We are currently introducing our services, which provide AI training (video learning + workshops) called "Corporate Reskilling®" and our in-house developed ChatGPT for businesses, "AI Works," to many companies, including listed companies.

Another advantage is that our services can be used at low cost by utilizing IT implementation subsidies and grants. We also undertake contract development of AI systems, and through training, tool provision, and development, we aim to promote the development of AI human resources and the use of AI in Japanese companies, thereby improving productivity throughout Japan.

◾️Overview of implementation support services

If you have any of the following issues, please contact us:

• How to introduce and utilize AI in internal operations
• AI training for employees and human resource development
• Promoting business efficiency using AI
• Support for improving AI literacy for executives and managers
• Consulting on the introduction of AI tools using subsidies