Satar ShamsiGoushki

I've built end-to-end machine learning and data engineering systems, including an LSTM forecasting platform for AAPL stock prices, PostgreSQL market-data pipelines, and document intelligence workflows.

At GoalEarn, I build modular ETL pipelines that acquire, validate, transform, and load data into PostgreSQL. I've implemented incremental loading, structured logging, automated tests, HashiCorp Vault credential management, and a machine learning inference API.

My forecasting work combines TensorFlow/Keras LSTMs, Optuna hyperparameter optimization, MLflow experiment tracking, FastAPI inference, Docker, GitHub Actions, and CI/CD. I also identified and fixed data leakage affecting validation results.

My graduate research at ITMO University covered machine learning, mathematical modeling, image processing, data analysis, and high-performance computing, including music generation, epidemic modeling, image classification, and segmentation.