One of the key promises of machine learning is reducing the latency between information and action, allowing machines to make instantaneous decisions on the latest data available to the organization. Doing this can be easier said than done, though. In this talk, I will provide an overview of the state of real-time ML in production, explore the real-world challenges organizations face when attempting to architect and deploy real-world ML systems, and share some of the ways leading organizations are overcoming these challenges using open source and commercial tools.
