Dmitrii Cherkashin

Achievements

• Forked and maintain an abandoned open-source Kafka Connect DynamoDB Streams connector in production (12 tables, 300+ shards); cut its checkpoint write load ~5.7× with zero throttling
• Scaled the CDC pipeline through an 80× surge in data volume after a major product launch
• Retired the legacy data lake after migrating its history into Iceberg

Responsibilities

• Own the company's data platform end to end: availability, quality, access speed and cost of data for analytics and health calculations
• Designed and built a data lakehouse on AWS from scratch: S3 + Apache Iceberg, Glue, Athena, Airflow
• CDC replication from Postgres and DynamoDB via Kafka into the lakehouse
• Data catalog and lineage in OpenMetadata; column-level access control with Lake Formation tags (PII/PHI); GDPR export and erasure pipelines

• Replication from IBM AS/400 into ClickHouse with data-quality reconciliation and recovery tooling

Achievements

• Reduced data processing time from 17 to 7 hours, raising SLA compliance from 70% to 95%
• Cut ClickHouse storage usage by 40%
• Improved p90 query response time from 20 seconds to under 1 second
• Mentored engineers and gave internal tech talks on data engineering

Responsibilities

• Designed and implemented data processing pipelines with Airflow, Trino and ClickHouse in Yandex Cloud
• Took part in the migration from a monolith to FastAPI microservices

Achievements

• Accelerated business report generation from 1.5 hours to 1–2 minutes

Responsibilities

• Hired and managed the DWH team from scratch; set the strategy, processes and deadlines for the DWH division
• Designed the Data Warehouse architecture with Data Vault
• Built ETL pipelines from MongoDB, Postgres and APIs into the warehouse with Airflow
• Built data synchronization between partner ERP systems and the PostgreSQL backend, keeping partner data consistent
• Introduced data quality checks and data governance

Achievements

• Designed self-service replication for “one-click” table onboarding

Responsibilities

• Built ETL pipelines from MySQL, ClickHouse, Kafka and APIs into the warehouse with Airflow
• Implemented online CDC replication MySQL → Kafka → Exasol
• Designed and implemented the Data Vault DDS loader with Airflow DAGs and operators

• Created the DWH data model for Delivery Club and the logical data model for Citymobil from scratch
• Built e-commerce user portrait reports in Tableau; supported Spark processing scripts
• Reduced time from task creation to production reporting from 3 weeks to 1 week

• Designed the integration receiving checks from the national payment card system
• Designed and prototyped Data Vault data models and data marts in SAP PowerDesigner
• Worked with Greenplum, Vertica, SAS DIS, Informatica PC and SAP BO; wrote BRDs and FSDs