Sergey G. Email & Phone Number

Key Projects and Achievements- Algorithm Optimization: Refactored a clustering algorithm using PySpark, Scipy, and Numpy, achieving a 20x improvement in runtime and a 5x reduction in resource usage, resulting in annual cost savings of over $600K.- AI Pipeline Enhancement: Streamlined the core AI pipeline with PySpark, leading to a 15x improvement in runtime and resource utilization, saving over $250K annually.- Security Product Implementation: Led the development and deployment of a near real-time endpoint misbehavior detection system using AWS (Lambda, SQS, S3), Polars, Pandas, PySpark, GBT, and Airflow.- Streaming Framework Development: Played a key role as a core developer in creating a new streaming framework, focusing on I/O components and performance under scale load using Streaming PySpark, Hudi, Kafka, Airflow, and Locust.- Initiated and led several smaller projects, including: - Developing a framework for ML logging - Enhancing SQL queries, improvements related to SQLAlchemy usage and DB performance - Creating a feature to compare network performance across customers

Lausanne Area, Switzerland

I have led and participated in various projects across different divisions, including Swiss Universal Bank, International Wealth Management, HR, and Compliance. My project portfolio includes:- Dashboard Design and Delivery: Developed and implemented dashboards to enhance data visualization and decision-making processes using Splunk - Client Segmentation: Executed client segmentation projects to identify and target key customer groups. - Content Analysis and Recommendations: Conducted content analysis and provided actionable recommendations to improve user engagement.- Churn Prediction: Developed models to predict customer churn, enabling proactive retention strategies.- Product Recommendations: Created product recommendation systems to enhance customer experience and drive sales.Technologies: Python, PySpark, Foundry, Pandas, Scikit-Learn, NumPy, SciPy, Kafka, Hive, Hadoop, Bash.Technologies: Python, Foundry, Pandas, Scikit-Learn, NumPy, SciPy, Scala, Kafka, Spark, Akka, Hive, MongoDB, Scala, Hadoop, Bash.

Lausanne Area, Switzerland

Data masking projectTasks: - ETL - Data processingTechnologies: Oracle PL/SQL, Informatica, Shell Scripting.

I served as a Research Assistant specializing in venture capital funds and investment estimation, focusing on:- Implementing advanced techniques to analyze investment patterns and inter-company relationships using graph models (centralities, eigenvectors, densities).- Utilizing technologies such as Python, NumPy, and NetworkX to develop and apply these analytical methodologies effectively.This role equipped me with valuable skills in data analysis, financial modeling, and Python programming, enhancing my ability to derive actionable insights from complex datasets in the field of venture capital and investments.

Lausanne, Switzerland

Thesis: Probabilistic Machine Learning applied to automated incident resolution at Credit SuisseThe incident management system at Credit Suisse handles thousands of tickets every day. The tickets range from the reporting of application and infrastructure failures to staff being locked out of their computers after forgetting their password. The tickets are written in natural language, generally in English or German. The project consists of threee tasks:- Provide suggested resolutions based on the description- Provide suggested cateforizations of the problem and resolution based on text and metadata - Identify pathways of problems and relation between problems and applications Dataset: 14.5M of documents, 13GB in dumpTechnologies: Python, NLTK, Pandas, Scikit-Learn, NumPy, SciPy, R

Baden, Switzerland

The project goal: to estimate the suitability of different protocols as a replacment of MMS (IEC61850) protocol. I set up and implemented several prototypes based on OPC UA, WSDL and XMPP, measured and compared the performance. The results will be used to set up a communication model that could maintain the performance in a large scale applications in substation automatisation.Technologies: C#, TCP/IP, XMPP, OPC UA, OPC, Web Services.

I implemented information extraction (physical addresses) and classification system (decision tree C4.5) for web texts. Web sites categorization system (TF-IDF, SVM)Technologies: C++, STL.

Tomsk

Private information security paperwork management. Technical requirement and detail design security personal information templates. Dongles, generation of key-pairs and making certificates, cryptography programes, users’ technical support.

Master In Computer Science, Internet Computing

Master Thesis: Probabilistic Machine Learning applied to automated incident resolution at Credit Suisse

Diploma In Computer Science With Honour, It Security