Senior Data Scientist / Machine learning engineer Resume

SUMMARY:

I'm a data scientist, PhD - degree holder in econometrics and math modeling, experienced in machine learning algorithms development, implementation and monitoring.
Developing recommendation system using Collaborative filtering - Item, User and Model (SVD decomposition) based approaches;
Building machine learning algorithms from scratch (without using built-in libraries on high-load env.) for CTR (Click-through-rate) and ITR (Install-through-rate) prediction;
Developing API service using Python bottle server to provide results in request-response mode;
Risk and debt collection scorecards building with regards to Basel II and FICO credit rating grades requirements;
Segmentation of client’s database by churn probability, probability of purchase, Life-time-value in order to define loyal clients and increase KPI of marketing events.

SKILL:

Database: SQL (Oracle, MS, Postgre) - 5 years;

Datebase: noSQL (MongoDB, Redis, Hive), Amazon S3 - 2 year;

Statistical packages: SAS (Guide, Miner) - 1 year, R (programming and Shiny app development) - 1 year, Python (SciPy, NumPy, SkLearn) - 3 years;

Web development: Python (Django), JavaScript, HTML, CSS - 2 year;

Programming: Python, R, Javascript, Java. Multilingual (Russian, Ukrainian - native, English - advance, French, Spanish - intermediate).

WORK EXPERIENCE:

Senior Data Scientist / Machine learning engineer

Confidential

Stack: Python, SQL, Redis, Amazon S3.

Responsibilities:

Development and implementation of business models - churn, LTV prediction, recommender systems, automation reports, API services for developed models.

Data Scientist / Machine learning engineer

Confidential

Stack: Python, R, SQL, MongoDB.

Responsibilities:

Data mining, machine learning algorithms development, data analysis, building predictive models, web scrapping and providing related consulting services.

Scorecards developer / Data Scientist

Confidential

Stack: SAS, SQL.

Responsibilities:

to maintain research and data analysis using data mining, machine learning technology, build predictive models and provide related consulting services.
My projects portfolio contains all types of scoring - application (including reject inference analysis), behavioral, collection and fraud, Call-Center working load modeling, clients’ segmentation on probability of purchase (credit card activation and usage), probability of default or probability of debt collection.

Lecturer (Associate Professor)

Confidential

Responsibilities:

The job entailed preparing and reading lectures and labs on economics and math cycle as: models and methods of optimization, econometrics, stochastic modeling, decision making, mathematical programming, information technologies in economics, programming, analysis of complex variable function.

Leading specialist

Confidential

Responsibilities:

Duties involved managing and updating monthly, quarterly, yearly reports and software.
Realized as Collaborative filtering - Item, User and Model (SVD decomposition) based approaches. Script in several Python modules to manage following process - connect to the Database, reading data, training the recommendations model, predicting the recommendation scores for all users and items, writing the prediction data to Redis DB;
Developed server-side script based on Python bottle module to read and parse the request, process the needed data and read corresponding recommendation from Redis DB, send response as the recommendation back to client.
Building machine learning algorithms from scratch (without using built-in libraries on high-load environment).
Developed online learning algorithms as a Python module for CTR (Click-through-rate) and ITR (Install-through-rate) prediction. Module read and parse the request from online auction system, update a model for requests with known outcome and predict a score for requests with unknown outcome, send the result back to system.
Develop adaptive price algorithm for online bidding auction.
Developed a Python script that takes predicted CTR and other parameters as input and returns recommended and minimum price to bid on auction. It was a second price type auction.
Risk and debt collection scorecards building with regards to Basel II and FICO credit rating grades requirements.
Developed script in R that realizes reading data from Database, preprocessing, training the model, scoring the set of users, writing data to the Database. Used different type out-of-the-box models - logistic regressions, svm, random forest, neural networks.
Segmentation of client’s database by churn probability, probability of purchase, Life-time-value in order to define loyal clients and increase KPI of marketing events.
Developed script in R that trains a model, predict corresponding target value for each user, making segmentation based on predicted score and business requirements.

We provide IT Staff Augmentation Services!

We'd love your feedback!

Resume Categories

Client Services

Job Seekers

Visa Sponsorship