GumGum Innovation & Technology Culture

One of the challenges our team faces is training a deep learning model with minimal biases. The goal is to discern between safe and threatening content such that brands are only aligned with brand suitable content. Our priority is to ensure that our data is diverse and representative resulting in a model that is more objective than subjective.

The Data Engineering team at GumGum is excited to be migrating our key systems to a new stack consisting of Astronomer AirFlow, DataBricks Spark streaming and Snowflake for warehousing. Our ultimate goal is to build a single data model and process to unify our batch and streaming pipelines, reducing cost and the time to implement new features