Data Scientist

World Business Lenders · El Salvador, Guatemala · posted today ago

Listing supplied by Himalayas. 101 Careers did not originate this post and applications are handled by the employer.

About this role

About World Business Lenders

At World Business Lenders (WBL), we provide flexible, short-term commercial loans backed by real estate to help small and medium-sized businesses across the United States — particularly those facing difficulties with traditional financing. We're a fast-moving, results-driven organization that takes security seriously as we continue to grow.
This is a Full-Time Independent Contractor role with working hours from 9:00 AM – 6:00 PM Eastern Standard Time, Monday through Friday. We request that all CVs be submitted in English.

This role sits within WBL's Data team and focuses on credit, loan performance, collateral, and recovery modeling, helping the business better understand and predict how loans and their underlying real estate collateral behave over their life cycle.
The work involves large, national real estate and lending datasets. This is real data at scale work, often involving millions of records and large, imperfect, real world files.
A growing part of this role is building predictive and forecasting models using historical, realized outcomes to estimate future loan performance, default and payoff behavior, collateral value changes, recovery timing, costs, and expected outcomes. This includes applying machine learning and statistical techniques where appropriate, not just descriptive or backward-looking analysis.
This role will contribute to the development and validation of WBL's internal loan valuation and risk models and comes with a high degree of ownership and autonomy as our data and modeling capabilities expand. .

Early on, the day starts by joining the daily BLV alignment meeting to stay current on the project's philosophy, direction, and where things currently stand.
Most of the day is spent hands-on with data: pulling and cleaning records from large national real estate and lending datasets, cross-checking fields across multiple sources, and investigating data quality issues as they come up. A meaningful part of this involves directing AI-assisted tools to run deep, multi-step analyses efficiently, while carefully validating their output rather than accepting it at face value.
Several days a week involve model-focused work: running back tests of internal valuation and risk models against real, historical outcomes, building or refining predictive/forecasting components, and running sensitivity analyses.
Regularly, the role also involves presenting and communicating analyses and findings clearly to the team and stakeholders, translating technical results into actionable takeaways

Analyze and validate internal credit, risk, valuation, and recovery models against real, historical loan and portfolio outcomes
Build predictive and forecasting models that estimate future loan performance, default or payoff timing, collateral value changes, recovery outcomes, costs, and timelines
Design and run backtests, sensitivity analyses, scenario comparisons, and time-based analyses across large historical datasets
Work with large, multi-source lending and real estate datasets to extract, clean, reconcile, and validate data for analysis and modeling
Identify the factors most predictive of loan performance, collateral outcomes, and realized recoveries, and build simple tools that let stakeholders explore model results and scenarios
Investigate and clearly document data quality issues, edge cases, model limitations, and inconsistencies found in large, real-world datasets
Translate analytical and modeling findings into clear, actionable recommendations for underwriting, credit, pricing, portfolio management, and risk management
Maintain clean, reproducible, well documented analytical and modeling work so it can be independently reviewed, validated, or extended by others.

Education:

Bachelor's or Master's degree in Data Science, Statistics, Mathematics, Finance, Economics, or a related field.

3 to 7 years of experience in data science, applied modeling, or quantitative analytics, ideally involving large, real-world datasets.

Background or strong interest in finance, banking, commercial lending, U.S. real estate, or property valuation.
Familiarity with commercial or small business lending is a plus.
Experience with predictive modeling, forecasting, survival/time-to-event analysis, or other time-based estimation problems.
Experience with credit risk, default, loss, recovery, model validation, or scenario/sensitivity analysis is a plus
Experience reconciling and cleaning data from multiple sources or systems.

Fast learner, intellectually curious, comfortable getting up to speed quickly in unfamiliar domains
Detail-oriented without losing sight of the bigger picture
Strong data communication skills, able to explain complex analytical and modeling findings clearly to both technical and non-technical audiences, including senior management
Comfortable working with real autonomy and ambiguity, and proactive about asking the right questions
Rigorous, skeptical mindset, doesn't take a number at face value without checking it independently.

Strong proficiency in Python and SQL
Experience with statistical modeling, machine learning, forecasting, or related quantitative techniques, including model evaluation and validation
Experience working with large datasets, including writing efficient, performance-conscious data processing code
Experience building simple, shareable analytical tools or dashboards is a plus