Ideally, a company should have a 1:1 personal conversation with each customer — the kind of delightful experience that happened in physical stores. AI makes this dream scalable again, and we don’t think chatbots alone are the solution [and some of us have worked on chatbots since 2002!]. Instead, we believe in empowering customer representatives and augmenting the customer experience.

This is made possible by recent research in diverse fields, including natural language understanding, generation and style transfer, end-to-end-dialog learning, speech processing and computer vision.

Creating or choosing a product requires a precise understanding of goals, trends and constraints. Giving feedback about ideas and even coming up with new ones is now within reach of machine learning algorithms.

Tomorrow’s creators and customers will interact with machines to guide their thought process and inspire them, whether they are prototyping a new model of car or trying to find the best outfit for the summer.

Leveraging new classes of models such as Generative Adversarial Networks and adapting them to your domain, we can help you take advantage of how AI will reshape creativity and perception.

A self-driving car is continuously evolving within an environment that is uncertain and constantly changing. The behavior cannot be scripted in advance, so in order to make the right decisions, the algorithm must build a dynamic representation of the world in real-time and predict how its actions will affect the future. Similar adaptive decision making is present in many sectors such as financial services, energy services logistics, or gaming.

If you have a case where thousands or millions of micro decisions must be made in a complex and interactive environment, we can help you support, improve and scale these systems thanks to recent developments in reinforcement learning.