Using Airflow To Power Machine Learning Pipelines at Optimove with Vasyl Vasyuta

Data orchestration and machine learning are shaping how organizations handle massive datasets and drive customer-focused strategies. Tools like Apache Airflow are central to this transformation. In this episode, Vasyl Vasyuta, R&D Team Leader at Optimove, joins us to discuss how his team leverages Airflow to optimize data processing, orchestrate machine learning models and create personalized customer experiences.

Key Takeaways:

• (01:59) Optimove tailors marketing notifications with personalized customer journeys.
• (04:25) Airflow orchestrates Snowflake procedures for massive datasets.
• (05:11) DAGs manage workflows with branching and replay plugins.
• (05:41) The “Joystick” plugin enables seamless data replays.
• (09:33) Airflow supports MLOps for customer data grouping.
• (11:15) Machine learning predicts customer behavior for better campaigns.
• (13:20) Thousands of DAGs run every five minutes for data processing.
• (15:36) Custom versioning allows rollbacks and gradual rollouts.
• (18:00) Airflow logs enhance operational observability.
• (23:00) DAG versioning in Airflow 3.0 could boost efficiency.

Resources Mentioned:

• Vasyl Vasyuta
• Optimove
• Apache Airflow
• Snowflake
• Datadog
• Apache Airflow Survey

Thanks for listening to “The Data Flowcast: Mastering Airflow for Data Engineering & AI.” If you enjoyed this episode, please leave a 5-star review to help get the word out about the show. And be sure to subscribe so you never miss any of the insightful conversations.