• Providing Machine Learning, Al and Data Analytics services for the clients.

• Building various microservices to orchestrate the workload and perform OCR.

• Healthcare project built to perform claim processing by utilizing LLM to analyse contracts and business logic to determine rules for estimating allowable payments.

• Served as the driving force behind risk model development and implementation, essential for loan approvals, provisions, RWA calculations, and monthly performance monitoring.

• Led the development and implementation of risk models for loan approvals, provisions, RWA calculations, and monthly performance monitoring, leveraging advanced techniques and state-of-the-art technologies.

• Developed predictive models for Probability of Default (PD) and Exposure at Default (EAD) estimation using Logistic Regression, Decision Trees, Random Forests and model ensembles, integrating these models into the IT infrastructure.

• Key player within the Belgrade Modeling Team, responsible for developing predictive models for loan approvals and monthly performance monitoring.

• Built predictive models for Behavioral and Application scorecards, IFRS 9 PD models by utilizing Python, SAS and R, from data gathering, analysis, modeling and productionalizing.

• Automated model creation processes, enhancing efficiency and readability and keeping abreast of industry best practices.

• Supply chain oriented product development including inventory planning & optimization, size-level planning, time series and demand forecasting.

• Development of a fully automated demand forecasting pipeline.

• Harnessing the power of state-of-the-art machine learning models like Temporal Fusion Transformers, N-BEATS and Temporal CNN networks.

• Time series forecasting of customer sales utilizing statistical and ML/Al models and estimating financial impact on a retail business.

• Built Dash app to display the results and allow business units to override the forecasts and insert their own predictions.

• Performed web scraping of all partner clinics websites and based on the information from their site, applied LLMs to perform clinic classification.

Developed a novel Churn model based on clients demographics and initial sign-up information and third party data.

Development and implementation of Churn and Customer Lifetime Value (CLV) models, from data ingestion to model development, deployment and performance tracking.

Improving the demand forecasting processes for selected regions by ingesting new features and improving existing forecasting pipelines.

• Improving the existing and creating new Power BI reports and visuals.

• Cleaning and reorganizing data sources and migrating PPT and excel reports.