How to build an AI-powered mobile ad platform

Meet Chang Kim, the VP of Engineering at MOLOCO. MOLOCO is focused on empowering businesses by helping them take a data-driven approach to their mobile advertising, and specializes in processing large amounts of data while leveraging machine learning technologies. Watch an exciting interview with Chang Kim as Max Saltonstall takes a deeper dive into MOLOCO’s scalable ad-serving cloud architecture. Chapters: 0:00 - Intro 0:33 - Who is Moloco? 1:10 - How does a mobile ad get to my phone? 5:56 - Neural Network models in use 6:45 - Getting the system to scale up for high demand 8:51 - Building the ad serving pipeline on Google Cloud 12:30 - Online service architecture 13:17 - Running the bid processor on Kubernetes 14:16 - Storing and using ad data for training and analysis 16:14 - Data architecture diagram 17:02 - Using Dataflow to handle moving data around 18:03 - Reasons for using BigTable 18:49 - Optimizing for cost control 20:36 - Why Moloco choose Google Cloud Architecting with Google Cloud playlist → ht