Developing and maintaining several applications related to the process of videos ingestion and encoding. Mostly backend development in Ruby on Rails.

Acting in the research, implementation and deployment of models that improve Globo`s videos platform. From computer vision and machine learning models applied in the automatic detection of opening and ending credits sequences on shows episodes, to bayesian optimization of the video player playback configuration parameters. Also responsible for the development of the data aggregation job (Spark/Scala) from which metrics regarding user behavior and player performance during video consumption are generated.

Involved in doing several exploratory analysis on data from Globo`s video player. Such analysis results would lead into major restructuring of the data pipeline model, in which I contributed by participating in the new model conception, evaluating data quality and implementing video player plugins (in JavaScript) responsible for data collection.