AI is becoming more prevalent for applications like smart medical devices, robotics, and financial trading, where predictions are made in near real-time. However, there are many complexities associated with managing high-frequency data, the AI modelling approach, and integrating the models in a streaming architecture.This talk will focus on building a system to address these challenges through:Using a stream processing framework that incorporates time-windowing and manages out-of-order data with Apache KafkaSynchronizing data and managing mathematical assumptions with signal processing techniquesIncorporating time requirements when choosing and implementing AI modelsUpdating and caching the models in the systemDuring the session, we will walk through the process of performing signal processing and time-alignment and designing and deploying a machine learning algorithm for streaming data.