 Employment rates among refugees are low worldwide, especially in the first years after arrival. We asked ourselves if there's a cost-efficient and effective way to change this with a solution that could be used worldwide. We looked into a lot of data of refugees that arrived or came to Switzerland and the United States in the last decades. And the results showed a very clear story. The place people are sent to matters for their integration. So the city or town someone goes to can really act as an on-ramp towards successful integration or an obstacle. And we are looking into this in more detail. We really want to set out and change the way this is done today with a simple but powerful idea because today refugees are allocated at random within countries. And what we want to do is really match people to places where they are most likely to find a job. The way we do this is that we developed an algorithm that uses existing data that we have on refugees and identifies the synergies between people's characteristics and places. With machine learning and optimal matching the algorithm is able to predict the probability of success so the chance is to find a job at each possible location based on trends and patterns in the historical data. The algorithm is also able to take into account real-world constraints such as available capacities at different locations or distribution keys that governments might have. What the algorithm then does is really identify the optimal place for a person or a family and then makes a recommendation what the best match would be and where someone is most likely to find a job. The algorithm that we developed is dynamic and can respond to different changes at each location and take that into account. So what we do is really human-centered artificial intelligence where the recommendation that the algorithm gives is meant to complement and not replace the decision-making of case workers and officials. We looked into data in Switzerland and the United States and ran back tests and the results were very staggering. We could show that employment rates could be improved by 40 to 70% in both countries with that matching technology. This is still theory that we are now putting into practice. The algorithm is currently tested in the real world. In partnership with the State Secretary for Migration, we're running a randomized controlled trial in Switzerland among a subset of asylum seekers that are currently arriving in Switzerland. What we're doing is that half of the people in the trial are allocated to cities and towns within Switzerland with the algorithm. The other half of the people are allocated to places within Switzerland at random the way it is done today. What we're basically doing is really use the power of data to learn from the past and improve the future and our goal really is to not only help refugees rebuild their lives but also help them increase their chances to unlock their potential to contribute to their new home countries.