Need for Speed: Machine Learning in the Era of Real-Time

The world is impatient! Everyone wants to be aware of everything everywhere, all at once. Growing demand for quick decision-making has propelled the fascinating world of real-time machine learning (RTML) into the spotlight.

Over the years, brilliant minds developed and refined a number of fantastic ML techniques. Yet, the impatient users are boohooing: we want our data now (read: low latency), and we like our data fresh (read: dynamic data adaptability), and we also want it to be cheap (read: utilize resources efficiently).

In this talk, we'll explore the evolution of solutions to these challenges and how ML systems are advancing toward online inference and continual learning.

By the end of our talk, attendees will leave with an understanding of the latest RTML techniques and the essential factors to consider when designing and implementing real-time machine-learning solutions.