2024.02.08

Maison AI – a Support Tool for the Fashion Industry that Aims to Automate Processes Using Artificial Intelligence

In recent years, amidst the launch of various generative AI services, Maison AI, developed by OpenFashion, stands out as an AI tool tailored for the fashion industry.

The beta version of this service was released in August 2023, providing a support tool for fashion companies that incorporates both text and image generative AI.

In addition, it includes an 'AI agent feature' where AI is tailored to professions such as 'fashion designers' and 'pattern designers', allowing for more accurate responses to specialized questions within these fields.

Currently, the number of workspaces (a space that users can freely create within Maison AI to share information among companies or teams) has exceeded 1,300. The tool is being adopted with the aim of improving work efficiency, generating new ideas, and creating new businesses.

We asked Toru Ueda, the CEO of the company, about the background behind Maison AI's development, its utilization, current challenges, and future plans.

Applying AI for the Fashion Industry

Could you first tell us what prompted you to start 'Maison AI'?

Personally, I have always had an interest in AI. Around the time I founded OpenFashion in 2014, I was involved in an AI startup called LeapMind as a UI designer. I've always had a positive perspective on the use of AI.

From that time, I considered creating a tool applying AI to the fashion industry. However, the technology was not yet mature, and there was a shortage of AI personnel.

In recent years, with the emergence of image generation AIs such as 'Stable Diffusion' and 'Midjourney', and the announcement of text generation AI 'ChatGPT', the use of AI for mass consumption has expanded rapidly. That's when I began developing 'Maison AI' as a tool that applies AI for the fashion industry.

As a fashion company, what were the key points you were interested in when it comes to utilizing generative AI?

Firstly, the use of AI for tasks such as creating sample designs, crafting catchphrases, creating product titles and descriptions for e-commerce sites were common amongst each company.

Furthermore, in the companies that have adopted it, especially younger employees, are actively using it by creating their prompts (questions or instructions inputted by the user for interaction with the AI).

Despite being in the beta version, there are cases where users have been able to reduce their workload by about 15 hours per month.

Challenges and Potential of "Maison AI"

What challenges have you encountered while implementing "Maison AI"?

As companies grow in size, post-introduction usage and "penetration" become challenges. It's said that the internal penetration rate of generative AI use is around 10-20% in companies that use it. We feel that the penetration rate of companies that have introduced our tool is about the same.

While younger employees tend to be savvy, there are many historical companies in the fashion industry, and there are senior employees who say that they might not be able to use it even if introduced into the company.