● Worked cross-functionally with ML researchers, analysts, and backend teams to deliver data workflow for anomaly detection models. ● Built and refactored modular, testable ETL pipelines using YAML and Python to ensure maintainability and scalability. ● Processed structured and semi-structured data using memory-efficient Python tools like Polars, DuckDB, and Apache Arrow. ● Conducted data exploration and visualization using SQL, Python, and Grafana, driving actionable insights and feature engineering. ● Gained hands-on experience in data drift detection, model lifecycle understanding, and time-series processing frameworks. PS: great employer, interesting tasks, but eventually whole department was laid off due to cost cutting.

Company decided to start the new data warehouse project, organizationally apart from the IT Department, under a brand new department of "Analysis and Data Management". So I got job among other five crew members, because: I have a good and diversity IT background. I was interested in data processing technologies. Responsibilities: ● Estimation of required capasities. ● Performing on-premise software installation and upgrade. ● Automation of routine system administration tasks. ● Code delivery automation. ● Fix related issues. ● Configuring of Zabbix monitoring. ● Python programming. ● Backups. Achievements: ● 3 envirements: sandbox, development, production. ● Deployment of 2 vertica clusters on 10 bare-metal servers, 5 of vertica databases. ● 3 home-made docker compose apps (inspired by puckel) for airflow sandbox, airflow developmet, airflow production. ● Docker compose apps for additional services like jupyterhub, nginx, Jenkins etc. ● Minio server pool of 3 bare-metal servers. ● Ansible playboks for the vertica and minio installations. ● Operational data store on minio shared between environments. ● Python class for maintaining data between stage on vertica and operational data store on minio. ● Jenkins jobs and gitlab pipelines for the code delivery. Zabbix monitoring for the Vertica. One more administrator/DevOps has been growed up.

It was short term and useful project. At the start: ● Informatica Product 360 had stopped working several time per day and had not monitored. Responsibilities: ● JVM (java Virtual Machine) tuning (JConsole, Java Mission Control). ● Setup of the logs transfer to Elastic (nxlog). ● Setup of the metricts transfer to NewRelic. Achievements: ● Performance of Informatica Product 360 had stabilized, performance metrics and logs became available for the staff.