Self-pour beer systems are an emerging market, and solutions are mostly in-bar novelties that involve heavily customized installations. However, Hop is a scalable, modular, distributed self-pour system designed to be installed anywhere beer is sold. After a user is registered in the HOP system, they can visit any HOP station, have their identity confirmed via facial recognition, and pour themselves a cold draft beer.
The HOP system leverages native iOS, AWS Lambda, and Elixir in the form of a Phoenix backend and Nerves firmware controlling beer dispensing and measurement hardware. Jeff's talk will cover the lean build process implemented in the development of this product, as well as an in-depth look at the current architecture from the hardware/firmware to the customer-facing iOS app to the web backend that brings it all together.
Jeff’s career has been focused on applying computation to engineering problems. The more he’s studied machine learning, the more he’s become motivated to understand how it can be applied effectively in systems that interact with the real world.
At Very, he brings his applied mathematics and machine learning knowledge to a wide array of problems and projects involving images, natural language, social graphs, temporal data, and geospatial data. His academic background in control systems and robotics helped him become the director of engineering at Very.