Arena careers

At Arena, we believe that the next leap forward in applied AI lies in active learning systems and techniques like reinforcement learning. These systems are ‘curious’ – they continuously try new things in the world, making them better at adapting to and operating in a fast-changing, complex world.

While most AI learns by looking at billions of examples of labeled data, like tagged faces, active learning systems improve through interactions with simulated environments – think Neo from The Matrix learning Kung Fu by fighting Morpheus over and over in a virtual dojo. Arena turns real world problems into simulations – into dojos, or video games to train young AI “minds”. Similar techniques have been used in research labs to beat humans in games, from Go to StarCraft to Dota.

We’re excited to bring this game-playing AI out of the lab and into the real world by creating simulations and writing software to let AI explore and interact with real environments.

Today, we are focused on pricing. We use active learning agents (e.g., Gaussian Processes, RL) to find the optimal products and discounts that a company should offer to a particular customer at a given time. Over the past year we have raced our agents against people and existing software, outperforming both, resulting in sizeable increases in revenue for our customers. We are now building a product to enable big and small consumer companies to deploy better product assortment & pricing strategies, automatically. Pricing is just the start. Our long-term mission is to create active learning agents that can learn to operate effectively and safely in any new, real-world environment.