Development of a chatbot for car sales and customer support Responsibilities: * Designing and building Redshift Warehouse and ETL processes (in a team). * Creating a Text-to-SQL engine for client queries to the Data Warehouse on Redshift. * Designing and implementing a production-grade chatbot architecture within a team of three. The system incorporates structured and unstructured business data, along with PII and SPII data filtering to comply with American and European data protection laws. * Developed a microservice-based chatbot application in Python, allowing data processing pipelines to be configured in a no-code/minimum-code style, primarily through YAML files. * Created guardrails and a PII detection service used not only by the chatbot but also as a filter for external API requests. * Created Document detection CV model that fits 130 Kb size (to operate on devices, with hard throughput constrains) Achievements: Successfully deployed a scalable chatbot application to production, improving data processing efficiency and ensuring compliance with data protection regulations. Improved system robustness by implementing automated quality measurement tools that align with business goals. Technologies: Python, AWS, Ray, Haystack, Langchain, Kubernetes, Weights & Biases, TensorFlow, MongoDB, PostgreSQL (Supabase)

LinkedIn Agent System May 2023 - Aug 2023 Writing agents for communication in LinkedIn Responsibilities: Designing and implementing a chat-bot system for communication in LinkedIn; Development of chat-agent systems, utilizing the OpenAI GPT3.5 API; Management of Kubernetes; Technologies: NDA Uniswap Analytics on GCP March 2023 - May 2023 Backend Programmer Writing an analytics service for evaluating the balance of cryptocurrency wallets on Uniswap. Responsibilities: Designing application architecture; Development, deployment, and support of services working with the BitQuery API (cryptocurrency transaction analysis); BigQuery database management; Development of analytical queries to BigQuery; Development of dashboards in Google Looker Data Studio. Achievements: Developed and delivered a data analytics project on the GCP platform. Technologies: Python, GraphQL, BitQuery, BigQuery, SQL, GCP MLOps in CV (Building Plans Recognition) June 2022 - March 2023 Building a production-ML pipeline in an RnD project for generating 3D models of buildings. Transitioning the project from an RnD stage to a Production stage. Responsibilities: Designing and implementing a machine learning pipeline in Computer Vision; Consolidating a large park of ML models and data processing stages into a single structure; Code refactoring, covering the DE part with tests; Writing project documentation; Achievements: More than doubling the speed of model development and testing processes; Together with the team, we went from disjointed communications and experiments to a scalable ML system. Technologies: Python, FastAPI, Apache Airflow, Kubernetes, OpenCV, Gitlab CI, GCP, ClearML

Building a system of recommendations, demand forecasting, optimizing a sets of promotions. Development of neural networks for text and visual search in the product catalog. Responsibilities: Collection, preparation of data; Model development and validation; Code review and documentation; Development and supporting of the code base; Writing tests for the code base; Clarification of customer requirements; Development planning, evaluation of product complexity; Participation in the development and implementation of the solution architecture; Key results: In all projects, achieved an increase in sales due to the modernization of existing approaches to the collection and intellectual processing of data. Created a CI process for machine learning models based on Jenkins, MLFlow on Kubernetes Delivered a production-ready reference project on the AWS platform and maintained it. Technology: Python, Pandas, Pytorch, Pytest, Amazon Web Services (AWS), Google Cloud Platform (GCP) Google Kubernetes Engine (GKE), Tableau, XGBoost, CatBoost, LightGBM

Work as an analyst and data scientist, partly as a data engineer (building ETL processes). As a data scientist, mainly worked with texts. Responsibilities: Study of the media field around the bank (text analysis from online media); Text analysis, web scraping; Text mining, web scraping Exploring processes inside the bank by information tails Analyzing middle-size textual data in order to provide meaningful representations for leadership Data-oriented code reviewing Scoring model validation Participating in solution architecture design and implementation Forecasting of a client's complaint, an employee's illness; Key results: Really helped the client to be understood and get feedback from the bank. My text clustering pipelines have been successfully used in various departments of the bank for a long time. Technology: Python, Pandas, Dask, PySpark, TensorFlow, NLTK, Sklearn, BeautifulSoup, ScraPy, Selenium

Supporting, debugging, fixing, extending functionality of the desktop application for medical service company. Interface test covering, writing backend for web services with the same to desktop functionality. Responsibilities: Unit test covering Debugging, fixing, extending desktop app functionality Writing backend for web services Writing, profiling SQL queries Refactoring custom ORM in project Key results: Increased the speed and memory efficiency of the ORM and optimizing key SQL queries to the DB. Technology: Python, Flask, Bottle, Unittest, PyQT

Responsibilities: Developing trading strategies based on deep supervised learning from different kind of data Developing life system with cryptocurrency exchanges integration Backtesting strategies Key results: Some strategies stable beated the BTC holding benchmark (proof of algorithmic trading with ML support concept on crypto market). Technologies: Python, Keras, Backtrader