Case Study
Tapjoy uses Amazon SageMaker to Deploy ML Models in Days Instead of Months
Tapjoy is a mobile advertising network. They bid into auctions and have to make split second decisions on how much they will pay for that ad.
Nick Reffitt, VP Data Science for Tapjoy, explains how they are using SageMaker to deploy ML models in days instead of months.
Tapjoy is a mobile advertising network, which has relationships with advertisers and publishers. They bid into auctions, where they have to make split second decisions on how much they will pay for that ad.
They have a round trip time of 300ms to respond back with bid, but only 5ms for the data science component.
Therefore it’s very important that they have an ML platform that can handle high volume and low latency. They chose SageMaker as it offers an end-to-end, feature rich ML service, is fully managed, and enables them to chop and change which ML frameworks are used.
It has enabled them to reduce the time it takes to train, build and deploy an ML platform from 3-6 months, to less then a week.
What is SageMaker
