Priyanka Beniwal

WORK HISTORY

EDUCATION

ABOUT PRIYANKA BENIWAL

I’m a Manager - Data Science & Analytics at TransUnion CIBIL, working at the intersection of credit bureaus, machine learning, and real-world lending decisions.I specialize in building scalable, bureau-driven models for the financial services ecosystem - from NTC risk models to industry-wide income estimation solutions used across lenders. My work spans the full lifecycle: feature engineering, model development, validation, deployment, and automation, with a strong focus on models that actually survive production.Over the years, I’ve designed 600+ predictive features on large consumer loan portfolios, working extensively with raw bureau data. I’ve led efforts around special value handling, missing data treatment, iterative feature selection, and stability monitoring, ensuring models remain robust across time and segments. I enjoy turning messy, high-volume credit data into signals lenders can trust.A key part of my work has been building automation-first analytics. I developed an R Shiny-based EDA platform that streamlines bivariate analysis, PDPs, interaction studies, and model comparisons - cutting analysis time significantly and improving consistency across teams.I’ve contributed to and led development of risk scorecards and generalized industry models, using algorithms such as XGBoost, GLM, and custom cGLM frameworks, and have hands-on experience with both consumer credit data and emerging commercial use cases.Technically, I work comfortably across Python, R, SQL, and big data ecosystems including PySpark, SparklyR, Hive, and Hadoop. What drives me most is building data science solutions that meaningfully improve and support risk management, underwriting decisions, and financial inclusion.