The fashion industry moves at a rapid pace, and brands are under pressure to stay ahead of trends. Designers often spend weeks on data analysis, trend forecasting, and creating countless sketches. This manual, often intuitive, process leads to wasted time, money, and materials. It also has a significant environmental footprint, with the textile industry causing around 1.2 billion tons of CO2 emissions annually. That’s the challenge German startup yoona.ai aims to solve with its intelligent fashion tech platform. Using artificial intelligence, yoona.ai enables brands to make data-backed design decisions in seconds, speeding up the design process and reducing waste from the outset.

This not only gave Michel access to a powerful support network to help the company grow, it also persuaded Michel that Google Cloud was the right provider for yoona.ai.

“The accelerator was the best we've ever done,” says Michel. “The support and the mentors were amazing. It brought us closer to Google Cloud and showed us it was not just a technology provider, but a team we could work with for long-term success.”

yoonai.ai migrated to Google Cloud in 2024. With credits to invest in its infrastructure, and the support of Google Cloud partner Seibert Group GmbH, the company built a modern, flexible architecture to manage everything from data storage to complex AI workloads.

For example, a designer can tell the agent to find the bestselling products from a previous season for a specific target market. The agent automatically analyzes the data and presents the designer with a report and even streams the information into an image generator to suggest new designs. This allows designers to spend less time on manual research and more time on creativity, resulting in designs that are more likely to sell.

Using such a large amount of compute power can be expensive. yoona.ai received a 20% discount on Google Cloud services, and it also worked closely with the Google Cloud team to understand how to use resources efficiently to optimize costs. Following the team’s advice, yoona.ai resolved to use powerful A100 GPUs for computationally intensive training, while using more cost-effective machines for day-to-day operations and inference.

“The A100 GPU is essential for what we do,” says Volker Ißbrücker, co-CEO and co-founder of yoona.ai. “Very few cloud platforms can provide these GPUs. Google Cloud gives us access to several hundred, meaning we can maintain our services even if something goes wrong. With the Google Cloud team we had the expertise to help us configure the system to run our models efficiently.”

Beyond compute power, yoona.ai is also very careful about data security, as its customers share highly sensitive proprietary information, including bestselling products and brand identity. With Cloud Storage, yoona.ai can ensure all that raw data is securely stored. Google Cloud’s commitment to European GDPR regulations and data sovereignty are also key to keeping designers’ intellectual property safe and secure, helping the company to build trust with customers.

By helping designers make smarter, data-backed decisions from the start of the creative process, yoona.ai is tackling the problem of waste head-on. The company calculates that by enabling designers to reduce the number of design sketches and product prototypes, it can help companies reduce carbon dioxide emissions in the development of their clothing collections by up to 93%.

As yoona.ai continues to train more specialized AI agents to meet the evolving needs of its customers and improve the fashion industry’s sustainability, it’s confident that with Google Cloud it has the reliable, scalable platform to expand its reach and help more fashion brands design the future.

“As a startup, before using Google Cloud, it was tough to manage our business model efficiently,” says Ißbrücker. “Google Cloud infrastructure, and the support of the team, have enabled us to reach many more customers.”

Michel agrees, “Our main goal is to provide the best service and make our customers happy. With Google Cloud, that’s exactly what we’re doing.”