Maksim Blekhshtein

Developed AI-powered resume ranking system using OpenAI API, and Python(Pandas, Requests, JSON) to automate candidate-job matching. Built a cloud-based BI platform on Google Cloud, integrating BigQuery, Apache Airflow (Cloud Composer), and PySpark for ETL automation. Designed Power BI & Looker Studio dashboards for real-time analytics. Optimized workflows, reducing manual coordination time by 5+ hours/month. Conducted data quality checks on 1,000+ transactions, resolving classification and reporting errors.

Developed a Python script using RADIS and NumPy to simulate and analyze infrared spectra, optimizing parameters to minimize deviation from experimental data. Refined error minimization by adjusting artificial spectrum parameters, using cost function optimization to enhance accuracy. Organized laboratory reactor setup, coordinating with management for equipment acquisition and software integration, and simulated reactor processes for operational optimization and visualization.

Built a Python interface to enable seamless communication between incompatible Fortran and C-based models, automating data transformation and preprocessing, which made previously impossible simulations executable. Developed a mathematical model for microwave emission in plasma and applied it in simulations. Collaborated with Swiss Plasma Center (EPFL, Switzerland) on two research projects, contributing to Python/C software development with external teams.

Developed a mathematical model for spectral analysis, optimizing parameters via gradient descent and minimizing a cost function with weighted regularization. Designed and implemented Gaussian fitting models for experimental data, optimizing parameters via numerical methods to improve signal analysis accuracy. Conducted Monte Carlo simulations to model physical processes and collaborated with developers to debug and optimize the simulation algorithm.

Enhanced and maintained software tools by updating Matlab, Linux, and Python packages for hardware control and data processing. Refined mathematical models for measured values conversion, improving the precision of beam composition measurements by 2-3 times. Developed and presented data-driven insights by improving Matlab-based data analysis tools and showcasing results to the Swiss Plasma Center team and management.