Asparuh Emilov

Results-driven Data Analyst with 4+ years of experience with a proven track record at Telus International Bulgaria and Fastmarkets, skilled in predictive modeling and analytical thinking. Enhanced team efficiency by automating data visualizations and streamlining processes, while fostering collaboration through clear communication. Committed to continuous improvement and operational excellence, leveraging strong problem-solving abilities to drive impactful insights.

100 Days of Code: The Complete Python Pro Bootcamp

• Used Python to develop a Churn Forecasting algorithm for the company with the potential to save couple of millions of Revenue loss each year.
• Created and deployed successfully end to end a webapp that is using CrewAI Agentic system and an ML trained model as a tool to predict income >=50K USD based on the famous adult dataset from the UCIMLRepo.
• Successfully automated WBR/MBR/QBR report that usually took an entire day of manual work to just 15 min.
• Developed a highly accurate multivariate time series forecasting meta-model in Python to reliably forecast expected tickets volume on an hourly basis for the entire month.
• Created my own local chatbot that can answer questions from my documents using RAG.

- Proficient in Python and SQL for analytics and

data engineering, using NumPy, SciPy, pandas,

scikit-learn, XGBoost, LightGBM, and

statsmodels; visualization with Matplotlib,

Seaborn, and Plotly.

- Built forecasting solutions across multivariate

time series using statistical methods and feature

engineering, as well as deep learning

(LSTM/sequence models) for hourly and

monthly horizons; experience with backtesting,

cross-validation, and model monitoring.

- Developed production-ready analytics

workflows and dashboards, and implemented

streamlined, ongoing data management

processes to ensure reliability and

maintainability.

- Financial domain experience: understanding

of the investment process and portfolio

management; researched, engineered, and

validated alpha signals; scaled analytics over

structured and unstructured data to extract

actionable insights.

- Containerized data and ML workloads with

Docker for reproducible environments and

portable deployment; collaborated across

teams to deliver robust, operational solutions.