Elena Sofina

Collected, processed, and cleaned multi-channel campaign data to segment users and support targeted strategy. Initiated an internal ML pipeline for audio quality analysis, combining a fine-tuned YAMNet embedding model with an XGBoost classifier and planning CNN-based multi-class defect detection.

Tested voice call quality across major international carriers by routing international traffic to evaluate transmission performance. Identified and categorized network defects (dead air, silence, drops, one-way audio) and submitted detailed bug reports with raw call logs for engineering fixes.

Translated official legal and administrative documents from Spanish into Russian under strict compliance frameworks at the Ministry of Internal Affairs (Izhevsk).

Independently managed end-to-end translation of a full-length academic book from Spanish into Russian while ensuring exact terminological consistency.

Provided real-time live interpretation across Spanish, English, and Russian during an international cultural event to facilitate coordination between diverse international groups.

Grade: Distinction

MSc (Distinction) in Artificial Intelligence and Data Science, completing a dissertation comparing full-image vs. single-cell unsupervised clustering for platelet morphology analysis. Built an automated computer vision pipeline with a U-Net (5-fold cross-validation) and a convolutional autoencoder for feature extraction and artifact isolation.

BA in Linguistics with a specialization in Romance Philology and coursework focused on linguistics, English, and Spanish.