Konstantin Tretyakov Email and Phone Number

- designed and implemented replacement of unmaintained outdated Hadoop + Informatica installation for main data ingestion with Python tooling, reducing full pipeline cycle time from ~90 minutes to <5 minutes and parsing failure rate from ~25% to ~2%. Internal data storage optimised from 27 TB of raw CSV and log files on HDFS to ~1,5 TB of processed parquet files in Minio S3. 
Altogether, those solutions have led to major reduction in the amount of necessary resources and freed up ~1 TB RAM, ~100 CPU cores and ~25 TB of disk space.- developed multiple solutions for automatisation of manual data integrations with external consumers, reducing operational routine load on the Data Department by ~10 man-hours a week.- introduced SRE approaches to data processing: Infrastructure As A Code, CI/CD and Configuration Management, monitoring, alerting and logging. That increased maintainability, reliability, and level of control over data processes, decreasing troubleshooting and incident recovery time from 1-3 days to <2 hours and time to change system state from 1 hour to 1-2 minutes.

- built and supported e-commerce web infrastructure, providing 99.99% availability of services. Implemented canary deployment with automated issue tracking, reducing the number of incidents in the production environment by ~30%, implemented zero downtime products updating.- mentored development team in terms of infrastructure, ci/cd, metrics, took part in code reviews, adopted some of the best DevOps practices, decreasing average TTM by up to 40%.- built development infrastructure with automatic dev-stand and stage-area deployment, autotest framework and notifications system, increasing team's development and testing speed reducing the amount of bugs reaching production services.

- brought DevOps practices to the development team, built full CI/CD, monitoring, logging and alerting systems from scratch, decreasing product delivery time to customer and increasing development process speed.- built a number of production solutions on the company's clients infrastructure, integrated with the company's central resources.- developed solutions of automated recovery in externally inaccessible client's infrastructure to reduce failures impact on business processes.- took part in development of ML systems for web-chat and knowledge base products.