Looking for a job

Armine Panosyan

@arminepanosyan

Message

Data Scientist at Plat.ai who developed fraud detection and credit risk models using machine learning.

Armenia

Message

At Plat.ai, I developed and evaluated an autoencoder-based model for transactional fraud detection, and built and compared models for credit default prediction. I also automated recurring reporting with Python, SQL, and Pandas.

I applied SHAP and feature importance analysis to interpret predictions and identify risk drivers. My work has covered the end-to-end machine learning workflow, and my MS thesis explored image super-resolution using edge-guided diffusion models.

Experience

Work history, roles, and key accomplishments

Current

Data Scientist

Current

Plat.ai

Nov 2019 - Present (7 years)

Developed and evaluated an Autoencoder-based anomaly detection model for transactional fraud detection. Built and compared XGBoost, CatBoost, and Logistic Regression models for credit default prediction, and automated reporting processes using Python, SQL, and Pandas.

PyTorch TensorFlow Autoencoder XGBoost CatBoost Logistic Regression Feature Engineering Hyperparameter Tuning SHAP Python SQL Pandas A B Testing

Teaching Assistant – Programming for Data Science

American University of Armenia

Feb 2022 - Jun 2022 (4 months)

Mentored students in Python programming, fundamental machine learning concepts, and debugging. Led group sessions and one-on-one deep dives to strengthen technical problem-solving skills.

Python Machine Learning Debugging Mentorship

Education

Degrees, certifications, and relevant coursework

Yerevan State University

Master of Science, Applied Statistics and Data Science

2022 - 2024

Activities and societies: Thesis: EdgeSR: Image Super-Resolution Using Edge-Guided Diffusion Models. Relevant Coursework: Deep Learning, Computer Vision, NLP, Machine Learning, Math for ML, Digital Signal Processing.

MS in Applied Statistics and Data Science. Thesis on EdgeSR: Image Super-Resolution Using Edge-Guided Diffusion Models.