Dmitrii Turusov Email and Phone Number

Subsidiary company of the biggest bank of Russia with over 1 mln leads monthly. It was founded in 2019 and now has millions dollars of gross-margin a month with positive cashflow• Engineered a Data Warehouse using Postgres (+cstore_fdw) +Clickhouse+Bigquery and Airflow, laying the foundation for efficient data processing• Successfully developed and maintained 30+ ETL pipelines facilitating seamless data exchange across diverse platforms, including MySQL, Google Bigquery, Rest API endpoints, Kafka, Clickhouse, S3, and Google Analytics• Engineered 50+ stored procedures and functions to establish data marts, implementrobust database monitoring, and enhance master data management• Fully responsible for DWH performance (Grafana and database stat monitoring, , optimizing of SQL queries)• Developed a Rest API using Python FastAPI, contributing to improved data accessibility.• Trained and implemented of machine learning models, including:– Similar-car model that increased average session lifetime by 4%, based on Item2Item ML algorithm the model determines similar cars and shows them to users for whom they may be interesting– Prediction of deal probability. It is used in backoffice to highlight for sale-managers those clients who have higher probability for buying a car. Another sphere of model’ application is to divide those clients with high propensity to buy and send them to manager with large background in car market. The clients with low propensity to buy, they are usually sent to front-line to communicate with call-center operators who are not very experienced in cars.

Startup company, that holds business in the sphere of C2B sales as cars. Daily it has about 10 000 auctions and 3000 sales with 2 mln dollars of gross-profit.• Developed DWH based on MS SQL Server and SSIS.• Developed 20+ data pipelines from diverse sources, overseeing the maintenance of 50+ pipelines to ensure data accuracy and reliability.• Designed and implemented web-scrapers in Python, to acquire data from car classifieds.• Developed and maintained Rest-API endpoints (C# webapi2).• Developed reports in PowerBI• Trained and implemented ML models:– Smart-margin ML-model which calculates the best price for every client on the base of information about car data (car brand-model-modification-production year-mileage-damages, ect) and client data (region,city, gender, age. time of visit at the point of sale,..) and calculates probability of event that a client sales his/her car. The result was +15-20% increase of gross-profit every month that was verified by A/B test.– Smart-slots project. Prioritization of clients to give forthose with higherrate-score more options of time slots to come to point of sale. After A/B test +7-8% increase of average gross-profit of clients visit.