Daniil Vyazhev

Currently:

• National Bureau of Credit Histories, Data Scientist ( From 2023-present )
• Computational Pragmatics Models and Methods Research Laboratory, Research Intern ( From 2024-present )

Work Experience:

ML and neural credit scoring models development team.

Area of responsibility:
Development of machine learning models.
Validation of machine learning models.
Conducting retro tests.
Monitoring of ML models.
Analyzing data.
Conducting A/B-tests.

Implementation and modeling experience:

Benchmarking product:
Developed a bank benchmarking system that integrates data, enables credit analysis and facilitates portfolio evaluation, loan default detection and debt analysis. Enabled institutions to comply and grow.

Automated ML Model Monitoring:
Developed a model monitoring dashboard to identify model performance and anomalies for prompt action, improving accuracy and decision making.

Automated Model Comparison:
Created a system to evaluate retro and product models, refining strategies using historical data to achieve optimal results and improve customer experience.

RNN models for predicting default and loan taking.
Various models of recourse network, gradient bousting, and recommender systems were developed for the tasks.