Shellpoint Mortgage Servicing Predictive Analytics Model Developer – Default Recovery in Tempe, Arizona
Who we are:
Shellpoint Mortgage Servicing (SMS) is America?s 5th-largest third-party mortgage servicing company. What is mortgage servicing? Our clients are businesses that own mortgage loans (such as banks and real estate investment firms); on their behalf, we manage (or "service") their loan portfolios: We collect homeowners? mortgage payments, pay homeowners? tax and insurance bills, and help homeowners in default to get current.
Customer and Business Focus. You love the process of analyzing and creating models and share our passion to do the right thing. You know at the end of the day it?s about making the right decision for our customers.
Curiosity. You stay on the leading edge of emerging data technologies and modeling methods and seek out opportunities to apply them. You thrive on diving into big data sets to answer questions and find behavioral trends. You are always asking questions and aren?t afraid to share your ideas. You develop a view on what?s happening in the marketplace but also listen to others.
Leadership. You challenge conventional thinking and work with stakeholders to identify and improve the status quo. You?re keen to share the insights you?ve gained with others and listen to their feedback.
Data-driven. You have hands-on experience developing robust statistical models using an array of modeling techniques on open-source tools and cloud computing platforms. You are about a lot more than model estimation statistics - you?ve carefully back-tested your models to know where they will perform well and where you?ll be looking to improve them in the future.
We are building our predictive analytics platform so we can help our customers make complex financial decisions and run our business efficiently. The predictive analytics model developer will combine strong data exploration, statistical modeling, communication and collaboration skills to deliver analysis and models that help the business understand borrower probability of default under various circumstances and optimize the opportunity to keep homeowners in their homes.
Become the subject matter expert on delinquency and default trends in our portfolio and across the US housing market. Do exploratory data analysis and collateral surveillance to understand emergent trends and factors driving delinquency.
Understand and use third party models used by the recovery business.
Work with a seasoned team of servicing and loan modification experts to create models which help our business identify opportunities and trends in collections and calling campaigns. Create models with an eye towards recalibration and adjustment for exogenous variables such as government housing policy changes.
Evaluate model tracking error, drill down into model weaknesses, understand model strengths.
Backtest your models and use subpopulation analysis to gain a deep understanding of model performance.
Deliver and deploy models into production environments; understand how users are leveraging your models.
Explain model analytics to business users and work with business to improve model capabilities; document your models for technical and non-technical audiences.
Do ad hoc data analysis to answer questions which emerge in a fast-moving borrower market
2-5 years of experience using Python, R, SQL, AWS to build econometric models
Familiarity with U.S. housing market, housing policy and mortgage products.
Knowledge of mortgage or other retail loan default modeling techniques
Familiarity with mortgage servicing rights, whole loans, and other consumer loan products
Familiarity with statistical regression techniques and predictive analytics methods
Ability to work with software developers, data engineers; understand the software development process and use reproducible research techniques
Excellent communication skills and ability to collaborate effectively with a team
- Advanced degree in a quantitative discipline such as financial engineering, math, statistics, science, quantitative economics.