Leverage the power of your data
Advanced Data Science firepower to industry
We use cutting-edge Artificial Intelligence algorithms to meet your business needs. Get ready to empower your data and boost your organization to the next level with our tailored Machine Learning solutions.
Focused on innovation
Dust AI is a research-oriented lab aiming to develop the highest technology. We seek to push the boundaries of knowledge and apply state of the art algorithms to deliver innovative solutions to our clients.
Artificial Intelligence for good
We believe that AI has the potential to solve society’s biggest challenges. For this reason, we collaborate with NGOs, committing to our mission of democratizing AI and having a positive impact in our community.
Our services
Data Analytics
Data-driven companies are more successful and profitable than their competitors. Using advanced data analysis and visualizations, we create big data solutions to help you identify new business opportunities, understand your clients, and find patterns and anomalies in your business processes.
Deep Learning
Using deep learning, you can extract the full potential of your data to automate any business process, forecast market trends, and provide added value to your services. We optimize your company by building intelligent algorithms that deal with text, images, audio signals, or any other type of data that you use.
Interpretable AI
You cannot depend on an AI system you do not understand. In our mission to guarantee trust and adoption, we deliver systems that are completely understandable by humans. With our interpretable models, we ensure that the model solves the right problem so you can focus on deriving value.
Check our latest projects
Regression Analysis for COVID-19
We analyzed the spread of COVID-19 in Europe during its initial phase, which allowed us to make estimations on the growth rate for different countries. We further deployed an accurate predictive model for the future number of deaths by leveraging the lag between reported cases and demises.
Robust Route Planner
Current route planners only consider shortest travel time, while neglecting the risk of missing any connection in the recommended route. We developed an algorithm that calculates optimal journeys between two points given an input risk threshold. We validated our system with Zurich's public transport network.
Intelligent Exercise Tracker
Using image analysis and machine learning techniques, we prototyped a software able to record a person, and identify and keep track of 10 different types of exercise. This product could be installed at gyms to give a personalized and intelligent exercise tracker service.
Detecting Astroturfing on Twitter
An astroturfing attack consists of the use of automated bots to boost the appearance of a particular keyword in the trending topics of social media. We presented an unsupervised algorithm to detect suspicious patterns in tweet publications based uniquely on time-series data.
Extreme Winds Risk Assessment
Extreme winds in dry regions are a major hazard, as they have an important role in spreading wildfires. We used extreme value theory to estimate the frequency of extreme winds throughout the year in Southern California. Moreover, we assessed the risk of forest fires in the region.
Meet our team
Pablo Cañas
Co-founder
MSc in Data Science at École Polytechnique Fédérale de Lausanne, BSc in Computer Science at Universidad Carlos III de Madrid. Pablo has experience in industry as a software engineer at Amazon and as a machine learning engineer at a fintech startup, and in top-tier research institutions, such as the University of California San Diego and CERN. He is specialized in deep learning, NLP and computer vision.
Andreu Arderiu
Co-founder
MSc in Statistics at École Polytechnique Fédérale de Lausanne (EPFL), Double BSc in Mathematics and Physics at Universitat Autònoma de Barcelona. Andreu has worked developing anomaly detection algorithms at Swisscom, as well as a researcher at Durham University and EPFL. He is specialized in mathematical and statistical modelling, and is currently doing research on causal inference and AI fairness.
