Custom software for optimal decision-making

Energy savings in institutional buildings

XWatts is an energy start-up that helps large institutional buildings reduce their energy consumption. We've helped them build software that calculates optimal schedules for the use of appliances using weather forecasts and temperature models. The software remotely manipulates devices installed in the buildings in real time.

Flight ranking

At OptionWay, we applied machine learning to historical data in order to model travelers' preferences. We  developed a ranking algorithm that selects the best (rather than the cheapest) flight to display on top of the search results.

Price optimization

At Expedia Partner Solutions, we applied machine learning to historical data in order to model price elasticity. This is used to help Expedia's partner agencies put together travel packages. The algorithm includes a strategy layer which lets the business define complex objectives (beyond just revenue maximization). The system processes millions of requests every day.

Route planning

We helped a travel start-up plan multi-day sightseeing iternararies for its travelers based on their perferences. We developed an algorithm that picks relevant attractions and schedules them in a way that minimizes transportation times and makes sure to alot enough time to each of them during up-to-date opening hours. The software was integrated to the customer-facing app.

Building footprint detection

We helped the French space agency detect buildings in images from Pleaides satellites, which are two satellites that together take a picture of the entire earth every day. We used machine learning coupled with an optimization algorithm to outline object footprints in the images.