Data Scientist - Risk Modelling

Basic Qualifications
● Masters or Bachelors degree in, Statistics, Economics, Machine Learning, Operations
Research, Computer Science or other quantitative fields. (If M.S. degree, a minimum of
1+ years of industry experience required and if Bachelor's degree, a minimum of 2+ years
of industry experience required)
● Proficiency in SQL and other analytical tools/scripting languages such as Python or R
● Deep understanding of statistical concepts including descriptive analysis, experimental
design and measurement, Bayesian statistics, confidence intervals, Probability
distributions
● Should have an understanding of defining and Testing of Hypothesis, statistical measure
of central tendency, Population and sample, sampling Techniques, Correlation and its
measures and CL theorem
● Proficiency with statistical and data mining techniques including generalized linear
model/regression, logistic regression, random forest, boosting, trees,dimensionality
reduction algorithms
● Proficiency with machine learning techniques such as clustering, decision tree learning,
naive Bayes
● Able to deal well with uncertainty and unstructured problems to be solved
● Should have an experience working with structured and unstructured data with applied
Data science/Machine learning techniques
● Define experiments based on the model to understand and thus enhance the model
● Experience in retail lending is a plus


Key Responsibilities
● Develop end to end Credit Risk scorecards ranging from applications to behaviour to
collections scorecard using techniques such as linear model/regression, logistic
regression, random forest, boosting/bagging trees, dimensionality reduction algorithms
● Optimize models’ outcomes to help business and drive growth
● Analyze, interpret and present outcome/results to stakeholders; Set up model monitoring
and understand the reasons of model expectations vs actual outcome
● Own and deliver multiple and complex analytic projects. This would require an
understanding of business context, conversion of business problems in modeling, and
implementing such solutions to create business value.