Vitalii Kotliarenko is enthusiastic about building distributed fault-tolerant systems and pipelines for big data processing. At Grammarly, he works in the Core Services team, which develops and supports back-end of the product helping millions worldwide to improve their writing.
This talk will put the light on challenges of building natural language processing pipelines at scale. Practical questions to be covered:
– quick infrastructure setup using AWS and Terraform
– rapid prototyping with Apache Zeppelin
– turning a proof of concept into the product
– testing Spark applications
– monitoring, performance optimization, and capacity planning
Code examples in Scala will demonstrate work with various Spark APIs.
All rights reserved JEEConf ©
