Zornitsa Tonozlieva

• • Developing ml models for electricity price forecasting in the Bulgarian energy market

• • Building system imbalance prediction models achieving 92%+ directional accuracy with confidence-based trade selection strategies

• • Forecasting energy consumption and production patterns to support grid balancing operations and costs

• • Working with time-series data pipelines using InfluxDB

Implemented predictive models utilizing Pandas, NumPy, and scikit-learn to identify market trends and opportunities. Developed LLM-powered documentation tool for regulatory model compliance, automating technical documentation generation. Leveraged AWS technologies (EMR clusters, S3, Athena, SageMaker), Hadoop, and Spark/PySpark to process and analyze large-scale financial datasets Authored complex SQL queries handling datasets exceeding 1 million records to generate automated reports that improved stakeholder visibility into key performance indicators. Architected back-end systems for an enterprise analytics platform enabling clients to leverage machine learning models for financial decision-making. Built robust data validation and processing pipelines using Pydantic, Pandera, and Jinja for complex code generation and templating.

Engineered data solutions in a fast-paced startup using Python (Pandas, NumPy) and AWS services (Lambda, CloudWatch, S3, DynamoDB). Developed backend systems, contributed to critical business decisions, authored technical design documents, and identified product enhancement opportunities through data-driven insights.