MLOps

ML-DRIVEN MEDIA RESEARCH

Customer
Leading Eastern Europe market player in media research, advertising monitoring and media (over 11 billion events in database in per day).

Challenge
Creation universal workspace for Data Scientists to scale the team and attract external Data Scientists to build ML models in the company area.

Project Results
As a result of the project, the company received a scalable and manageable space for the development of ML models, which allows you to quickly engage internal and vendor Data Science teams.

Business Value
Significantly reduced the costs to create and implement ML models into a production environment. Decreased time-to-market for new analytical products based on machine learning models.

Tech Stack
Kubernetes, Jenkins, Jupyter, Mlflow, Feast, Tensorflow, Airflow, Spark Operator, Seldon Core, Git, Java, Apache Kafka.