Smriti Karki

About Smriti Karki

I am a results-driven Data Scientist with a strong foundation in machine learning…

Experience

1. Data Scientist

   AT&T

   Feb 2025 — Present · NY, US

   Contributed to the development and deployment of machine learning models supporting customer churn prediction and fraud detection, helping improve retention strategy and personalized marketing initiatives.• Performed exploratory data analysis (EDA) on large-scale telecom datasets including call detail records (CDRs) and mobile telemetry using Python and SQL, uncovering key usage trends and behavioural patterns.• Collaborated with data engineers to build real-time ETL pipelines using PySpark and Airflow, enabling automated processing of CDR and IoT sensor logs for downstream modeling and analytics.• Supported dashboard development in Power BI and Tableau, delivering executive reports on KPIs such as churn rate, ARPU, and customer lifetime value.• Contributed to the design of A/B tests for campaign optimization and onboarding flows, analyzing results with statistical methods including t-tests and effect size interpretation.• Participated in Agile sprints and worked cross-functionally to prioritize use cases in customer segmentation, network usage forecasting, and NLP-driven ticket classification using SpaCy.• Assisted in deploying models using Docker and tracking performance with MLflow, ensuring reproducibility and basic CI/CD integration on Azure.• Documented workflows, findings, and model assumptions in Confluence, promoting transparency across data and product teams.