Karol Ďuriš
Verified Expert in Engineering
Data Scientist and Developer
Zalužice, Košice Region, Slovakia
Toptal member since February 25, 2026
Karol is a data scientist with 7+ years of experience delivering ML solutions across fintech, banking, travel, and industrial domains. With a focus on predictive modeling, customer analytics, and data-driven decision support, Karol has proven results such as doubling ancillary sales and improving hiring success rates.
Portfolio
Experience
- Statistics - 15 years
- Applied Mathematics - 15 years
- SQL - 11 years
- Pandas - 7 years
- Jupyter Notebook - 7 years
- Machine Learning - 7 years
- Python - 7 years
- R - 3 years
Preferred Environment
Windows, Jupyter Notebook, Cursor AI, Slack
The most amazing...
...thing I’ve created is a recruitment scoring model that improved hiring success from 22% to 46% using candidate data.
Work Experience
Data Scientist
Vacuumlabs
- Performed feature engineering and selection across large datasets (100s of features) to improve model robustness and interpretability.
- Led a 2-person data science team to develop predictive models for detecting wound coils improperly on industrial TVA winding machines.
- Developed a predictive scoring model that increased internal recruiting success rate from 22% to 46%.
- Developed ML modules to enhance SME/micro-credit scoring and credit-limit setting processes.
- Developed transaction categorization logic for Indian retail banking data, enabling automated classification across multiple expense categories.
- Extracted and structured financial advisor client data from PDF statements using automated data-processing pipelines.
- Designed and developed a company-wide reporting system consolidating financial, employee, and client data into a centralized analytical platform.
Data Scientist
auxmoney
- Contributed to the development of price elasticity models to quantify customer sensitivity to interest rates and predict conversion under different pricing scenarios.
- Built ML models to predict loan conversion probability using partner behavioral and financial data.
- Designed and automated monitoring of key pricing KPIs, ensuring model stability and early detection of performance drift.
- Conducted ad-hoc analyses that identified key profitability drivers.
- Developed price elasticity models to quantify customer sensitivity to interest rates and predict conversion under different pricing scenarios.
- Built and improved regression-based models for pricing decisions, incorporating partner data and behavioral features.
Data Scientist
Datatree
- Developed algorithms to assess clients’ financial health, defining and implementing key metrics such as financial reserve and cash-flow stability indicators.
- Designed and validated outputs of an automated financial advisory system, ensuring accuracy and reliability of personalized recommendations.
- Built transaction categorization logic for card payments in the absence of MCC codes and implemented bank transfer classification by detecting recurring payment patterns and behavioral signals.
Data Analyst
National Bank of Slovakia
- Developed macroeconomic analyses and medium-term forecasts to assess fiscal sustainability and public finance risks.
- Analyzed long-term sustainability of the Slovak pension system, incorporating Eurostat demographic projections and aging population dynamics.
- Processed and integrated data from multiple economic and administrative sources to support macroeconomic analyses and public finance forecasting.
Experience
LLM-powered Investment Assistant for Portfolio Insights
The system leveraged LangChain to orchestrate API calls for retrieving market data and financial news, enabling context-aware responses grounded in real-time information. I implemented query classification and parameter extraction (e.g., time horizon, assets of interest) to dynamically route requests and retrieve relevant data.
The assistant combined structured financial data with unstructured news signals to generate meaningful, user-friendly explanations of portfolio movements.
ML-Ancillary Revenue Optimization
ML-based Candidate Scoring System
The model increased recruiting effectiveness from 22% to 46% by prioritizing high-potential candidates and identifying key attributes associated with hiring success. The system improved sourcing efficiency and enabled more data-driven screening decisions.
Loan Pricing & Price Elasticity Modeling
Industrial Defect Prediction
Automated Financial Health Scoring & Advisory Platform
Education
Master's Degree in Informatics and Applied Mathematics
Univerzita Komenského v Bratislave - Bratislava, Slovakia
Bachelor's Degree in Informatics and Applied Mathematics
Univerzita Komenského v Bratislave - Bratislava, Slovakia
Skills
Libraries/APIs
Pandas, Scikit-learn, Beautiful Soup, XGBoost, Imbalanced-learn, NumPy, Matplotlib, PyTorch
Tools
Git, Slack, PyCharm, MATLAB, n8n, BigQuery, Claude Code, GIS, ARIMA
Storage
Databases, MySQL
Languages
SQL, Python, R, C++, Snowflake
Frameworks
LightGBM
Platforms
Jupyter Notebook, Windows, Google Cloud Platform (GCP), Docker, Vertex AI, Databricks, Kubeflow
Paradigms
Anomaly Detection, ETL
Other
Statistics, Data Science, Linear Regression, Regression Modeling, Statistical Analysis, Data Analysis, Data Analytics, Data Cleaning, Data Modeling, Data Preprocessing, Datasets, Analytical Thinking, Analysis, Mathematics, Mathematical Finance, Machine Learning, Applied Mathematics, Risk Models, Risk Modeling, Data-informed Recommendations, Forecasting, Probabilistic Modeling, Logistic Regression, Feature Engineering, Pricing Elasticity, Pricing Models, Time Series, Data Engineering, Sales Forecasting, Model Evaluation, Data Labeling, Statistical Learning, Statistical Modeling, Classification, Algorithms, Large Data Sets, Credit Risk, Cursor AI, Economics, Econometrics, Statistical Methods, Predictive Modeling, A/B Testing, Customer Lifetime Value (CLV), Financial Modeling, Recommendation Systems, Web Scraping, Website Data Scraping, Large Language Models (LLMs), Artificial Intelligence (AI), AI Agents, AI Model Training, Time Series Forecasting, LangChain, Agentic AI, Anthropic, Prompt Engineering, Vector Databases, API Integration, APIs, Time Series Analysis, Machine Learning Operations (MLOps), ETL Pipelines, GeoPandas, Natural Language Processing (NLP), Actuarial, Documentation
How to Work with Toptal
Toptal matches you directly with global industry experts from our network in hours—not weeks or months.
Share your needs
Choose your talent
Start your risk-free talent trial
Top talent is in high demand.
Start hiring