Nikita Denisov

I am the Team Lead of International Ecom Search Team in the International Search Department @ Yandex and have 4 people under my command. Me and my team are creating search, ranking and recsys solutions in Turkye in the following places: - https://yandex.com.tr/search/?text=iphone+17&lr=11508 (products widget "Mağazalardaki fiyatlar" or "Talep üzerine popüler ürünler «iphone»" found in here: https://yandex.com.tr/search/?text=iphone&lr=11508) - https://yandex.com.tr/shopping/ (shopping vertical: search, ranking and recsys) Responsibilities: - managing the team of 4 people - establishing strategical and development plans - reporting directly to the CTO, product leads and key managers - improving quality of search via developing new models, conducting feature engineering and creating new offline and online metrics for better understaning of different aspects of search

At this role I am responsible for leading products search and ranking stream in Yandex International Search for Turkye. Resposibilities: - Ownership and leading of major project - Reporting directly to the CTO of international search in Yandex and to the lead product manager of this stream - Establishing strategic plans - Managing and assigning tasks to the employees under my command - Cross-team collaboration with backend development team and QA testing team - Mentorship and leadership of colleagues under my command Achievements: - Conducted 2 consecutive successful A/B tests that helped achieve half-year plans in one quarted - Reduced amount of irrelevant documents in SERPs in 4 times - Created quality metrics with respect to the team's OKRs - Achieved smooth transition from one backed system to another as a result of cross-team collaboration and strategic planning of all the details - Developed the pipeline for data preparation and model training and deployment - Collected golden set for offline validation of our improvements

As and RnD ML Engineer in Search&Ranking i was responsilbe for the end-to-end development and integration of ML solutions into the product Resposibilities: - identifying and formalizing business need - conducting feature engineering and experiments using ClearML - trainig various ranking models - integrating new features and models into production via ML Flow and runtime code written on Go - conducting A/B tests and deep analysis - cross-team communication with product analysts, QA and backend developers - mentoring newbies - conducting interviews for middle and senior positions Key achievments: - implemented various CTR-based features than showed a +2% increase in NMV (net analog for GMV) as well as growth of orders by around +5% - implemented runtime features for prices, availability etc that resulted in +1.5% increase in NMV and +3% in orders - developed unified excperiments pipeline (data preparation, models traing and metrics calculation) that significantly decreased time need for offine experiments - refactored key features and data pipelines (code, CI/CD)

I work in a recommender systems team at Alibaba Group (AliExpress Russia). I create recommender algorithms for different shelves of application. That includes: creation of most suitable recall for each shelf, training ranking models, analysis of metrics, further improvement of recommender systems, deployment to production. I have strong experince with Python, Apache ecosystem (Airflow, PySpark, Hadoop, HDFS), MLFLow, ML, DL, Linux, Bash. I implemented complementary items recommendations algorithm based on the ArXiv article. I developed a pipeline for next click prediction using Bert for Masked Language Modelling based on another ArXiv article. I started a BI service for visualization (ReDash) on remote cluster for our team, so that we can monitor various metrics of our data and pipelines. I created an inference pipeline for neural networks in PySpark. I took part in improvement of our team's service for evaluation of recommender algorithms, that included work with front-end (Kotlin) and back-end (FastAPI). I constantly take part in analysis and research of possible improvements for our product.

At VTB Bank I worked in Collection Department in team of analysts. I was responsible for establishment of monthly plans of collection for various loan portfolios, I developed a visualisation system for these plans (to make establishment as pipeline: SQL -> Analysis using Python -> Visualisation using Python, HTML and CSS -> Final negotiation of established plans). I created several ML models for prediction of loan payments for different groups of clients. I had experience mentoring younger colleagues and presenting to upper management.