How Otrium uses artificial intelligence for robotics and content creation

Artificial intelligence (AI) is rapidly conquering the fashion industry. While one company is still waiting, another is already moving forward.

Dutch outlet platform Otrium uses artificial intelligence (AI) both behind the scenes in the warehouse and at the forefront of generative AI content. Co-founder Milan Daniels explains how this will make the company faster, smarter and more customer-focused. He also discusses the opportunities he sees for the future.

How do you see AI? What does Otrium use artificial intelligence for?

As a company, we are excited about the possibilities that AI offers us. Developments are progressing quickly. For us, it’s about testing, learning and scaling what works.

We currently use AI in two ways: In our warehouse we use robotics and automation, including ‘Autostore’. This is an automated pick-and-pack machine that uses AI to calculate the fastest routes to assemble orders. This saves costs and makes us faster for our customers.

Additionally, we use generative AI to provide images for products for which we do not receive product or model photographs. This improves presentation and increases conversion. We’ve been doing this for about four months. We introduced robotics and AI in our logistics process at the beginning of this year in April.

How did you come to implement automation and AI in your warehouse?

Implementing automation and AI in our warehouse is technically very demanding and requires collaboration. We implemented this together with our Dutch fulfillment partner Bleckmann, the Swiss company Kardex, which manages the machines, and the Norwegian company Autostore itself. All parties worked together to prepare for the installation. There is decades of experience behind this, as Kardex and Autostore also do this for many other companies.

For us it was mainly a question of the right time to start. It was a significant investment. It was only possible because we have now reached a scale where it makes economic sense. That was a milestone for us.

What impact did the AI-controlled warehouse have on your operations? What did it achieve?

The difference is enormous. We have more capacity and need less staff as the workforce has been reduced by 70 percent. It’s also good that the warehouse employees have fewer repetitive tasks. As already mentioned, we can serve customers faster.

Added to this are the efficiency gains. We invest the money saved in better service or in our marketing budget. This brings us closer to profitability.

In addition to an ordering machine, we also have Genesis, a machine that packs boxes. A polybagging machine will also be added soon, which will automatically pack products into a plastic bag.

At the same time, each system has its own learning curve. Working with people presents challenges, and the same goes for robots. Sometimes they stall and need to be restarted. This vulnerability and technical complexity are also part of it.

How do you use Generative Artificial Intelligence (GenAI) and what benefits has it brought?

Otrium works with 250 brands and receives five million items annually. Sometimes model photos are missing or there are no usage rights for third-party platforms like ours. Since we primarily work with remaining inventory, the depth per SKU (editor’s note: inventory unit) is often low. This has previously made it unprofitable to organize expensive photo shoots. However, a product will not sell well without model photos. Thanks to GenAI we can offer this cost-effectively and thus improve the presentation.

We now also create inspiration images and campaigns internally, for example for targeted newsletters. While traditional fashion brands work with seasons, we are dealing with constantly changing products. GenAI allows us to quickly create content around them. We would never have been able to do it to this extent manually. The customers’ reactions to this were also very positive.

What does AI require of your employees? GenAI also brings its own challenges…

The challenge with GenAI is that new tools and services are constantly coming onto the market. This creates a lot of movement and a certain amount of chaos. You might invest in a solution and a month later a new version or alternative comes out. This works ten times better and is part of a larger suite, like Google Gemini.

This requires enormous flexibility from our team. It means not committing to one solution, but constantly re-evaluating which tool best suits what we want to create. This is really different than before, where you would choose a SaaS (Software as a Service) solution and work with it for years.

Where do you see the greatest opportunities for AI in the future?

There are some directions in which we see a lot of potential. Sizingas the size issue, is a very important aspect. Of course not just for us, but for the entire fashion industry. With the help of AI, we can move towards greater consistency in sizing and potentially learn from returns to give customers better sizing advice.

A second area is moving images. The model photography is now static. But if it becomes videos, it offers new functionalities where you can rotate or zoom. This brings the online experience a few steps closer to that of physical shopping.

Do you have any advice for other fashion companies regarding AI?

I hate to tell others what to do, but I can share our experiences. Otrium started early with AI lookbooks, even if they didn’t look that convincing at the time. We knew things would get better quickly.

My takeaway is: Don’t wait until something is perfect. Start early with AI and grow as it develops. Anyone who joins later risks losing touch and missing out on a competitive advantage.

Otrium recently launched a new segment, Beauty. The company is also active again in the UK, with the market now served directly from its main warehouse in Almelo, the Netherlands.

Does your company use AI in design, logistics or inventory management? Contact us at [email protected] and share your experiences with the editorial team.