Utilize advanced mathematical, statistical, and analytical expertise to research, collect, analyze, and interpret large datasets from internal and external sources to provide insight and develop data driven solutions across the company
Build and test predictive models including but not limited to credit risk, fraud, response, and offer acceptance propensity
Responsible for the development, testing, validation, tracking, and performance enhancement of statistical models and other BI reporting tools leading to new innovative origination strategies within marketing, sales, finance, and underwriting
Leverage advanced analytics to develop innovative portfolio surveillance solutions to track and forecast loan losses, that influence key business decisions related to pricing optimization, credit policy and overall profitability strategy
Use decision science methodologies and advanced data visualization techniques to implement creative automation solutions within the organization
Initiate and lead analysis to bring actionable insights to all areas of the business including marketing, sales, collections, and credit decisioning
Develop and refine unit economics models to enable marketing and credit decisions
What you’ll need:
5 to 8 years of experience in data science or a related role with a focus on Python programming and ML models.
Proficient in Python programming and libraries such as NumPy, Pandas, Scikit-learn, TensorFlow, Keras, PyTorch.
Strong understanding of machine learning algorithms, deep learning techniques, and natural language processing methodologies.
Strong Python/Pyspark knowledge enabling data preprocessing and historical feedback loop for context loading for LLM.
Familiarity with SQL databases (MySQL/ SQL Server) and vector databases like (Qdrant, Faiss).
Proven track record of delivering successful AI projects that demonstrate measurable impact on business outcomes.
Strong analytical skills with the ability to interpret complex data sets.
Master’s or PhD in Computer Science, Data Science or related field preferred