• Implemented robust and scalable backend services for Samsung TV and phone apps.

• Designed and led backend services for Samsung’s Highlights app that includes microservices for the mobile app and CMS, video encoding serverless system, data ingestion, infrastructure, and CI/CD (NestJs/Typescript, AWS, Terraform, k8s).

• Implemented and owned one of the backend services of the cloud gaming platform that allows end users to play games via TV without gaming consoles (NestJs/Typescript, AWS, Terraform, k8s).

• Participated in building a backend service to support NFTs in the TV app (NestJs/Typescript, AWS, Terraform, k8s).

• Helped other teams with infrastructure and CI/CD for their projects (AWS, Terraform, k8s).

• Implemented the Video Annotation and Training AI Pipeline (GCP, NodeJS, ReactJS, Python, Kubernetes)

• Implemented the Video On Demand service: upload and stream videos (GCP: GCS, Transcoder API, BigQuery, GKE; NodeJS, Kubernetes)

• Designed and implemented the Data Platform on GCP for the Data Engineering team (Pub/Sub, GCS, BigQuery, Dataproc/Spark, Composer, Python, Terraform)

• Designed the new version of the Anki Data Platform.

• Implemented the Anki Data Lake (AWS S3, Glue Data Catalog, Spark/Parquet) and access to it (AWS Athena, Redshift Spectrum, EMR/Spark).

• Added EMR/Spark as a processing engine (running on demand by Airflow) to the Data Platform to unload resource-consuming parts of ETL from Redshift.

• Designed and implemented ETL/data modeling/data quality for a new product, robot Vector, to support product analytics and business needs.

• Migrated the legacy ETL for robot Cozmo to use the new EMR/Spark/Airflow framework.

• ETL - create automate pipelines data processing and monitoring (Bash, Python, Airflow); extract from different sources (mostly APIs), transform and load data to DB (Amazon Redshift).

• Data modeling – involved in some data modeling (e.g. SCD type2) for some new tables for existent DWH.

• Data Quality Assurance – create Data Quality Check scripts for data profiling to discover inconsistencies and other anomalies in the data to improve the data quality and email alerts if issues with data exist.