Ivan Petrov

Experience

• 

  MBTS-Bank

  Nov 2007 - Apr 2008

  PL/SQL Developer

  Programming reports in Diasoft db, supported managers in getting reports on bank branches performance:• Developed reports to control limits/incomes in branch offices• Developed report for synchronizing Diasoft data with cards owners db

• 

  NPF Lukoil Garant

  Aug 2008 - Apr 2009

  Actuary

  Worked with pension methodologies and actuary calculations:- Actuary calculations – pension estimation, editing Pension rules- Consulting programmers on actuary calculations in Caruso software- SQL+ queries in Oracle db – optimizing process of data collection for daily reports

• 

  MEC

  Jan 2010 - Mar 2016

  Data Scientist in Media

  Data analysis and research methodologies development in the Quantitative Marketing field; leading software and tools development; lead group of four people (2 analytics, 1 db admin, 1 C# programmer); analyze, read and search science literature to adapt new research methodologies.Software Development:o SQL scripts for processing advertising datao R - Sites grabber using RSeleniumo R – ETL optimization, econometrics models, visualizationo R + RSelenium served on AWS Unix – web based tools for paid search budget estimationo Python – programs that work with Yandex, Google and VK APIs; site scrapperso Python + PyQt – developed stand alone applications for internet buzz analyses using IQ’Buzz API;o Parallel programming in Python – optimization performance for apps that utilize external APIsData Science:● Statistical experience – network visualization; cluster analysis using k-means, Bayesian classifier, EM classifier; Principal Component, Multidimensional scaling, correlation analysis● Econometric/Machine Learning models – marketing-mix models, calculating ROI, optimizing media budget. Multidimensional time series models, panel models, ordered and binominal regressions, PLS-SEM modelling, Attraction models. Optimization using different computational algorithms: partial least square technique, OLS, WLS, maximum likelihood, Ridge regression, LOESSo IKEA traffic modelling and media budget optimization, traffic predictiono Conducted model that predicts traffic in 300 Eldorado stores; Modelling OOH effect on each store depending on distance between billboard and storeo PLS models for calculating effect of advertising on brand KPIs; sponsorship evaluation; awareness modellingo TV + Digital data fusion based on monte carlo simulations for binomial modelling of advertising effect on brand awareness; Modelling effect of different number of ads/spots that were seen by user on brand KPIs.o Cox-Hazard model for digital attribution and conversion path analysis Show less

• 

  Onefactor

  Mar 2016 - Dec 2017

  GIS Data Science Specialist

  Company website: https://1f.ai/Company products: banking scoring, retail location scoring, marketing modelling Software Development:- R, Python- Git- Hadoop, HDFS- Spark, Impala, Hive- PostgreSQL, Hbase, Teradata, Parquet raw data- Linux, bash- Xgboost, NLP for clusterization- Kibana- Confluence for documentationData Science:Geo data mining using R and Python. DW and query engines: Spark, Impala, Teradata, Hive. Building future product prototypes primary based on GIS data from telecom users tracks:• Programmatic Digital Outdoor Billboards proof of concept: optimization spot scheduling to gain reach/affinity optimized plan;• Modelling store visits and building predictive model that proposes new store locations based on users tracks and mobile activity;• Clustering users tracks: intercity users travels into trains, planes, cars and etc.;• Optimizing and modifying outdoor placement plan for generating additional traffic to the retailer stores;• Classifying locations: searching for unload sites and cargo warehouses for trucks Show less

• 

  Sberbank

  Jan 2018 - Jul 2018

  Data Scientist at Wealth Management Department

  Company website: https://sberbank.ruCompany products: banking investment products, pension plans, insurancesSoftware Development:- Python- SQL TeradataData Science:Creating next product to buy models for wealth management department:• Predicting future users of individual investment accounts• Predicting buyers of home insurance• Clustering clients based on their attitude to usage of individual pension plans

• 

  McKinsey & Company

  Aug 2018 - Dec 2023

  Tools:- Python, Git, Spark, AWS, Azure, Gurobi, PytorchWorked in R&D group which was researching application of fine-tuning large language models for client specific tasks:o Trained NLP models on GPUs on cloud (GCP Vertex AI, AWS Sagemaker).Created digital marketing personalization engine for large retailer:o Customized first level ALS matrix factorization recommendation engine.o Recommendations for 10m+ customers on Azure cloud stack.o Implemented CI/CD using Azure DevOps pipelines, Databricks workflows and kedro library.o Optimized Spark queries that speed up calculations from several hours to several minutes.Developed rail freight delivery optimization model for a wagon park owner. Client has 60k wagon park and more than 150k deliveries every month. Task was to increase rail carriages utilization while minimizing cost: o Developed from scratch integer linear programming model using Gurobi package.o Achieved improvement in wagons park utilization from 87% to 94%.Created model for SKUs pricing elasticity estimation for a big retailer. Client has hundreds of stores and tens of thousands of different SKUs. Task was to achieve better pricing for non-KVI products:o Multiple time series forecasting of product store sales.o Did A/B tests with covariates adjustments.o Solution showed $10m+ improvement in sales if using optimal price policy.Applied classical Machine learning for different stages of oil refinery process. Worked with Fluid Catalyst Crackers and Hydro Catalyst Crackers:o Usually, you have 6-8 time series models for each of the factory final products.o On top of the product models, I’d created revenue optimizer.o Potential increment in revenues was up to $15m. Show less

  • Data Science Specialist

    Feb 2023 - Dec 2023

  • Data Science Specialist

    Mar 2022 - Jan 2023

  • Data Science Specialist

    Aug 2018 - Feb 2022