Mykola Tkachenko

Experience

• 

  NATIONAL RATING AGENCY RURIK

  Jan 2013 - now

  • Designed, developed and deployed the automated reporting system which allowed to reduce report preparation time by 75%:- Extraction & transformation of file data using Pandas; - Loading client information from a MongoDB database;- Loading transformed json data to the Ubuntu-based FastAPI application to generate comprehensive rating reports and analytical reviews using JinjaTemplates.• Designed, developed and deployed a back testing system (FastAPI application) to optimize parameters for trading strategies which allowed to daily assess the historical profitability:- Extraction and aggregation of minutely trade data from exchange catalogues and APIs using Pandas, Requests and Parquet;- Generation and optimization of trading strategy parameters using Optuna and BayesOptimisation. • Designed, developed and deployed the machine learning pipeline for a probability of default (PD) model for Ukrainian banks which allowed to significantly reduce the time of the quarterly parameter updates:- Extraction and aggregation of financial and rating data using Pandas and web scraping to build a learning dataset;- Training and serializing models using SkLearn and Pickle;- Loading best-performing models to build and visualize PD curves using Matplotlib. Show less

  • Chief Information Officer

    Aug 2024 - now

  • Python Developer / Lead financial analyst

    Aug 2018 - Aug 2024

  • Financial Analyst

    Jan 2013 - Aug 2018