Krish Sonani

Data Scientist @Bank of America

TX, US

Engineered predictive lending and credit risk models using Python, Scikit-learn, XGBoost, and Hugging Face, improving underwriting accuracy across loan accounts and reducing loan default probability through sentiment extraction. • Orchestrated end-to-end data ingestion workflows with Azure Data Factory, Synapse, Kafka, and Airflow, automating structured and unstructured data pipelines in Azure and reducing reporting latency by around 50 hours per month. • Implemented segmentation and recommendation models through K-Means, Regression, and Predictive Analytics, increasing targeted cross-sell revenue by $2.5M and improving customer retention across multiple retail segments. • Constructed interactive dashboards using Power BI, Azure Data Lake, and Matplotlib, providing treasury leadership with insights from PostgreSQL and MongoDB that accelerated capital reallocation decisions by $10M. • Collaborated under Agile (Scrum), developing secure ETL pipelines in Azure Databricks and leveraging FactSet, reducing risk data reconciliation errors by 15% across more than 1,500 credit exposure reports.

AURO University, Surat

Bachelor of Business Administration - BBA

The University of Texas at Dallas

Master of Science - MS