Skip to content
Customer story

Food Tech Curbs Spoils By Up to 70% with Automated Machine Learning Devices

Orbisk reduces food waste with a mere scan; using automated machine learning achieves success with global expansion plans ahead

Related services
Cloud, Data and AI, Machine learning

Industry
Travel / Hospitality

Company
g-company

The Utrecht-headquartered food tech startup Orbisk uses innovative technology and solutions to reduce food waste in the hospitality industry. The company was created in 2019 after founder Olaf van der Veen noticed how much food was wasted at a hotel breakfast buffet. Orbisk employs cutting-edge technology to scan leftover food and track and monitor waste. These plug-and-play devices are able to reduce food waste by up to 70%. The company recently secured a €2.4 million grant from the European Innovation Council allowing for international scaling and more accurate AI models. By 2025, Orbisk aims to eliminate 100 million kgs of food wasted in professional kitchens annually.

Why
Achieve scalability on an international level
What
Make operational machine workflows
How
Fully automate the machine learning architecture

An Appetite for Precision

Orbisk relies heavily on AI, machine learning, and innovative tech for its software to work. Images and weight of excess food are uploaded to the cloud where an AI scans and analyzes them at pixel level for ingredients. This worked well on a smaller stage, but Orbisk needed to scale on an international level. Each step of the process had to be automated and refined. The software needed to be finetuned. Orbisk asked Xebias experts to provide insight. They would help lay out the blueprint for implementing deep learning and machine workflows. The aggregated data and algorithms are then used to uncover key insights, creating more precision and efficiency to drive decision making.

Strategic Implementation of AI

Machine learning was key to Orbisk’s quick and successful deployment in the Netherlands. But to keep up with future growth and constantly improve, they needed to dive deeper into the possibilities of AI and machine learning. Employing Google’s Cloud was a logical next step for the Google Cloud for Startups alumni. As CTO Bart van Arnhem noted, “Tracking food waste needs to be as automated and seamless as possible.” The company made it a priority to liaise with Xebia to enhance their IT and machine learning environment using Google Cloud. Xebia’s expertise and guidance worked as a great sounding board for Orbisk’s own team.

“We have a quite technical team ourselves, of course, but we do get a lot of value from a partner like Xebia helping us review our architecture, and the decisions we make, especially for the longer term.”
Bart van Arnhem CTO Orbisk

Constant Innovation Fosters Growth

Creating the ultimate network for reducing food waste required great minds. The collaboration with Xebia fostered an environment of shared knowledge with both teams coming up with solutions. As Johanna Schacht, Orbisk’s team lead of artificial intelligence noted: “During one of the sessions, we spent eight hours asking questions back and forth. That was enormously valuable.” This constant back and forth led to finding that the best way forward was through automating the ML workflow entirely with well-kept quality control mechanisms in place. Automated infrastructure and data combined with streamlined machine workflows allowed Orbisk to focus on further research and development, growth, and innovation.

Explore related customer stories