Adel Kaliyeva

@adel-kaliyeva

Junior Data Scientist | Time Series Forecasting and Anomaly Detection

I'm working on air pollution forecasting, including models that predict up to 24 hours ahead. I also built a SIEM prototype for spotting anomalies in logs, so I'm always glad to chat about practical ML projects.

Career

Experience: 0 years

Experience

Data Engineering Intern

Itransition

Itransition is a global technology services company providing software development and consulting for businesses across industries.

Apr 2026 — Jun 2026

Built a Microsoft Fabric ETL pipeline and Power BI dashboards to integrate data and examine links between transportation, environmental, and economic indicators.

• Developed an ETL pipeline in Microsoft Fabric using the Medallion architecture (Bronze to Silver to Gold) to integrate data from multiple sources. • Built interactive Power BI dashboards to analyze relationships between transportation, environmental, and economic indicators.

Technical Support Intern

Damu Information Technology Center

The center supports DamuMed, a medical information system, and provides technical and information support to its users.

Mar 2025 — May 2025

Supported DamuMed users, handled requests, explained system functions, and maintained data in Excel.

• Provided technical and information support to DamuMed users, handled support requests, and explained system functionality. • Entered, validated, corrected, and updated data in Microsoft Excel.

Education

Master's, Applied Data Analytics

Astana IT University

2027

Bachelor's, Software Engineering

Astana IT University

2025

Selected projects

Air Pollution Forecasting

Performed data preprocessing, EDA, and feature engineering. Trained and compared ARIMA, SARIMA, Random Forest, XGBoost, and LSTM models. Co-authored an IEEE publication on air pollution forecasting. Currently developing a hybrid model based on LSTM, DFA, and the Generalized Cauchy Model for forecasting horizons of up to 24 hours. Benchmarking its performance against multi-output CatBoost and ne...

AEGIS: Anomaly Detection in SIEM

Evaluated CNN, LSTM, Isolation Forest, and Autoencoder models on HDFS logs using precision, recall, and F1-score, and integrated the selected ML model into a SIEM prototype. Set up log collection and processing, developed a FastAPI backend and Angular frontend, implemented Telegram alerts, containerized components with Docker, and deployed the prototype on AWS. In 2025, the project placed 2nd a...

Open project

Achievements

2nd place, university stage of the Student Research Competition (NIRS)

2025

3rd place, national stage of the Student Research Competition (NIRS)

2025

Skills

SQLPythonpandasNumPyExcelPower BIapplied mathematicsstatistical analysisdata cleaningexploratory data analysis (EDA)MLDLNLPbasic knowledge of LLMs and RAGTensorFlowscikit-learnmodel training and evaluationtime series forecastinganomaly detectionGit