— improving ClickHouse part of the storage by using various techniques — providing internal consulting on internals/peculiarities working with ClickHouse— facilitating the “client” side of the database (i.e. application layer) to achieve the best performanceand workaround weak sides of the clickhouse database subsystem

— introduced ClickHouse to the data science team, resulted in speeding up the Metabase dashboards (more than x10 in some cases). Successfully set up replication from postgres to clickhouse using pg2ch tool. — leaded the project of substituting postgresql with clickhouse for one of the biggest internal components, resulted in proposing better and easier to maintain architecture and a working prototype. — did internal consulting on how to work with Clickhouse or/and Postgresql: sharing best practices of using each technology— author of the continuous data stream tool from PostgreSQL to ClickHouse, talk on ClickHouse meetup in Madrid: https://murat.kabilov.com/pub/clickhouse_madrid.pdfvideo: https://www.youtube.com/watch?v=8gghC3OJYuU— participated in upgrading stack of the postgresql instances to the latest major versions with minimum downtime — participated in tuning postgresql configuration parameters to match type of the workload and fetch as much as possible from the hardware— enhanced postgresql metrics collecting system (author of https://github.com/adjust/postgresql_exporter/ )— author of the continuous backup solution based on postgresql logical replication ( https://github.com/mkabilov/logical_backup )— contributing to open source projects which company is using (mainly written in Go)— migrated postgresql database from amazon rds to the datacenter with almost no downtime.

- Started PostgreSQL Kubernetes operator project, as a hack-week project ( https://github.com/zalando-incubator/postgres-operator ), talk on pg day russia '17 ( https://murat.kabilov.com/pub/pg_kubed.pdf )- Provided internal consulting on database-related questions, mostly query optimization

— Maintained a huge amount of the legacy code written in PHP along with PL/pgSQL code (as for the stored procedures).— Have designed and implemented company-wide SQL Code Style for the databases and server-side developers.— Participated in constant review of SQL and PHP code— Played a significant role in training SQL and PHP developers to deal with the peculiarities of working with PostgreSQL— Job-interviewed a huge amount of the candidates following the principle "it's better to talk to a lot of people rather than miss a gem"— Enhanced website’s backend, e.g. made some parts of the system work asynchronously using self-written pgq consumer in php.— Have developed database architecture for side-projects like bekam.com, avito.ma (currently sold to the competitors). Have developed DB scheme which can be used to create new multi- language classified websites.— Completely remastered key part of the website regarding value added services pricing. — Developed database architecture for internal context advert system «Avito Context»— Proficient with profiling highloaded PG instances— Implemented PG Stored Procedure versioning and deploy mechanism— Involved in building development environment using Vagrant and Chef. I was alone responsible for building copy of all the database infrastructure (including obfuscated test data) inside Vagrant.