YES!Delft

Ubiops: Platform for AI and ML models

Published
Many scientists are great at writing code for Artificial Intelligence (AI) and Machine Learning (ML) models, even though it’s not easy. However, another big challenge is to create a custom application to run those models in the cloud.

Usually, this takes much time to build. To make this easier, UbiOps offers data scientists and engineers a platform that helps them run, manage and scale their models, saving them the trouble of building and maintaining such infrastructure themselves.

The road to Ubiops

Yannick Maltha met Victor Pereboom during his chairman year at YES!Delft students. They became friends and often talked about entrepreneurship. Then, seven years ago, helping BAM with some important data challenges, they decided to take the leap and start their first business: Dutch Analytics. One year later, Jorick Naber also joined as co-founder. Over the years, their activities evolved from helping customers build AI and ML models to a platform to deploy those models efficiently. That is why they began to turn Dutch Analytics into UbiOps three years ago.

Yannick Maltha: “When building AI and ML models for customers such as Prorail, Siemens, BAM, and FrieslandCampina, we learned many things about the requirements for turning models into scalable cloud applications. With this experience, we started building the UbiOps platform.”

When building AI and ML models for customers such as Prorail, Siemens, BAM, and FrieslandCampina, we learned many things about the requirements for turning models into scalable cloud applications.Yannick Maltha, co-founder and CEO of UbiOps

Making a difference

The target market of UbiOps consists of companies that want to productize their data science: building digital products from code and scaling them. The users of their platform are usually data scientists or engineers. They can use UbiOps to manage their code in the cloud without building a custom environment, and the platform also allows running multiple models simultaneously.

UbiOps offers its platform to multinationals as well as startups. Their services provide special discounts for early-stage and later-stage startups to support them in applying innovative ideas and scaling up faster. They do so because UbiOps believes in helping startups to spread their innovative ideas. They help both early-stage and later-stage startups. For example, they work with DuckDuckGoose, which specializes in deep fake detection.

Yannick Maltha: “We have a customer creating AI models for creating more effective immunotherapy to treat cancer. They use UbiOps to run their models. It is rewarding to see customers with these kinds of products advance using our platform.”

We have a customer creating AI models for more effective immunotherapy to treat cancer. It is rewarding to see such customers advance using our platform.Yannick Maltha, co-founder and CEO of UbiOps

The journey of Ubiops

UbiOps followed the validation program of YES!Delft. They significantly benefit from the network they have gained through the incubator. There are startups within YES!Delft using UbiOps, which has helped to validate the product. So far, they have been funded by BAM, the Rabobank, and Global Founders Capital.

Currently, they are in the scaleup phase. The platform is mature, helps dozens of customers, and has over a thousand users. They are active in multiple countries, with employees in the UK and USA. The aim is to expand the team, and they work together with YES!Talents to recruit new employees. 

Ubiops logo and platform UI showing on a laptop screen

One of the critical keys is to scale while keeping in mind data privacy. Many European customers using UbiOps are building fundamental AI and have strict regulations concerning storing their data in the cloud. The other important factor is sustainability: the calculation power of the GPUs used generates CO2 waste and heat. UbiOps is already working with parties to ensure residual heat is used as district heating. 

In the upcoming year, UbiOps wants to scale using partners in the ecosystem, both with consultants and cloud providers. To grow, they want to expand their marketing and sales internationally. For example, they are at the World AI Summit in Montréal, Canada, in April 2023. In short, they are ready to tell the world about their platform and UbiOps, so they can help accelerate AI and ML technology.

Your dream can become a reality too!

As long as you want to make a change, don’t miss the opportunity to boost your startup by joining the YES!Delft Validation Lab or Accelerator Program. However, if working for a startup resonates more with your ambition, we got you! Check out vacancies placed by our startups.

Be the change, as change starts with you.

To top
1
2
3
4
5
6
7
8
9
10
11
12