Kickstart AI Challenges You
to Make an Impact

Our objective is to accelerate the adoption of AI and champion it as a powerful tool, contributing to collective growth and prosperity. Our Challenges light a fire under the AI ecosystem in the Netherlands — creating an environment that sparks ideas, connects the people who can make things happen, and leads naturally to innovation. The solutions we build are then handed to our Product Team, who develop them into open-source products to scale-up our impact and enable more people benefit from AI.
‍
‍Partner Challenges are focused collaborations with our founding partners that allow us to collectively address business challenges and strengthen our community.
‍
‍Academic Challenges bridge the gap between academia and industry, giving insight and momentum to the next generation of AI experts.
‍
‍Public Challenges extend our impact further, with participants from diverse backgrounds joining forces to learn from one another and do amazing things with AI.

‍

💡 Want to see Kickstart AI Challenges in action? Read about our past projects:

Using Reinforcement Learning to Further Slash Food Waste

Kickstart AI x Ahold Delhaize Food Waste Challenge 2.0
‍After last time's success we decided to do another food waste challenge! This time we help Albert Heijn to tackle food waste on the items that need it most; the easily perishable ones. From vegetables to fresh sushi and pizza. Our goal is to have everything sold at the end of the day while having no customer face an empty shelf.During this challenge we optimize Albert Heijns discount mechanism and make it dynamic using reinforcement learning. We use a variety of variables to predict customer demand throughout the day and learn what that means for different products and how that should affect discounts. To really push Albert Heijn forward on this topic we got a dream team of top data scientists and AI experts from Ahold Delhaize, ING, KLM, NS and Kickstart AI.

Using Data to Prevent Food Insecurity

Kickstart AI x Ahold Delhaize Food Waste Challenge
Food waste is a huge issue in the Netherlands and across the world. A big part of the problem is the difficulty of predicting customer demand — there are so many factors at play, it’s hard for retailers to know what to stock when, and how much of each product they need on the shelves. And getting it wrong doesn’t just cost the retailer, it harms the environment by wasting natural resources too.

So in 2022 we pulled together a dream team of top data scientists from Ahold Delhaize, ING, KLM and NS to get a grip on the problem. They worked together with Delhaize Belgium to analyse the data and explore how different factors could be modelled to produce more accurate predictions. Following three sprint weeks of development and months of in-store testing, we succeeded in improving customer demand forecasting by 26%. Ultimately this has reduced food waste at Delhaize Belgium by a whopping 21%! This was an amazing result which can potentially be applied across all of Ahold Delhaize’s brands. Nice work, team!

Leveraging Data from Sewage Water to Protect Communities

Kickstart AI vs COVID-19
The COVID-19 pandemic has had a massive effect on our everyday lives. The pressure on hospitals and medical professionals reached incredible proportions. Lockdowns created a struggle for individuals and businesses, everyone was doing what they could to maintain some semblance of normality. The worst is hopefully over now, but the risk of another outbreak still lingers.

Luckily, during the height of the pandemic a team of heroic data scientists and software engineers from Ahold Delhaize, ING, KLM, Philips and NS came together to fight back. Using measurements of viral load in sewage water samples from around the Netherlands, they developed a model to predict major coronavirus outbreaks before they could spread. The model was developed into a production-ready and platform-agnostic early warning system and handed over to the Landelijke Meldkamer Samenwerking (LMS) and the Instituut Fysieke Veiligheid (IFV), where it can make a real impact preventing future outbreaks.

TRAYS, Using Historical Data to Prevent Aircraft Food Waste

Kickstart AI x KLM
KLM has significantly curtailed in-flight food wastage, thanks to the implementation of artificial intelligence. TRAYS, KLM's newest AI model, which was launched at the end of last year by Kickstart AI, stands out as the inaugural model crafted exclusively for the airline's catering operations. Drawing insights from historical data, this AI model forecasts the number of passengers expected on board.

The AI model's predictive process initiates 17 days prior to departure, persisting until just 20 minutes before takeoff. This meticulous timeline ensures the generation of the most precise passenger numbers, guiding the entire catering process—from procurement to loading—toward preventing any excess of meals. This enables precise calculations for the required number of meals, resulting in KLM achieving a substantial 63% reduction in food waste based on the anticipated number of passengers per flight. Annually, this translates into saving over 100,000 kg of meals.

Kickstart AI Challenges You to Make an Impact

Kickstart AI Challenges You
to Make an Impact