Principal - Data Science

Work Location(s)2911 Lake Vista Drive, TX

Work Schedule:

Ally designates roles as (1) fully on-site, (2) hybrid, or (3) fully remote. Hybrid roles are generally expected to be in the office a certain number of days per week as indicated by your manager. Your hiring manager will discuss this role's specific work requirements with you during the hiring process. All work requirements are subject to change at any time based on leader discretion and/or business need.###

The Opportunity We are seeking a highly independent Data Scientist who thrives on taking end-to-end ownership – from problem framing and data acquisition to modeling, deployment, and impact measurement. This person will be an individual contributor focused on supporting Consumer Auto Servicing. This person will work closely with stakeholders to define high-impact opportunities, design scalable solutions, and ship models and analyses that surface insights and drive business strategy.

This person will also mentor developing analytics professionals and connect with other analytics and data science teams to build capabilities. This role is ideal for someone who can navigate ambiguity, consistently deliver high-quality results, and effectively work with cross-functional partners.
* Partner with strategy, operations, and business leaders to transform ambiguous questions into data-driven solutions that drive measurable outcomes.
* Communicate findings and recommendations clearly to technical and non-technical audiences and influence decision-making at multiple levels by making holistic recommendations with considerations for risks, tradeoffs, and blind spots.
* Design, execute, and interpret experiments and analyses to evaluate business strategy changes and quantify impact.
* Build robust, scalable models and analytical solutions; document thoroughly; and implement automated monitoring aligned with Model Risk Management standards.
* Benchmark challenger models and perform rigorous validation (back testing, stability, fairness, and drift) with monitoring and clear acceptance criteria.
* Productionize models and analytics pipelines in collaboration with data and implementation teams, ensuring reliability, observability, and reproducibility.
* Build reliable, documented data pipelines and features from enterprise sources; transform and curate datasets to support analysis, ad-hoc exploration, and model development.
* Develop high-quality data assets: feature stores, standardized metrics definitions, actionable dashboards, and monitoring for model and data quality.
* Own end-to-end project delivery: define problem statements, create project plans, align stakeholders, manage risks, and deliver outcomes on time.
* Manage projects of various size/complexity, day-to-day tasks, and priorities independently.
* Create strong partnerships with key partners, understand organizational goals, and shape roadmaps to support achieving them.
* Establish and promote best practices in modeling, code quality, and experimentation; mentor via code reviews and knowledge sharing (without direct reports).
* Proactively identify and learn new data sources, tools, methodologies, and approaches to improve performance and accelerate delivery.### The Skills You Bring
* Significant experience (5+ years) in analytics, data science, machine learning, or applied statistics with a track record of independently delivering production-impacting work and measurable business impact.
* Domain experience in financial services (preferably auto lending) focused on data/statistical analysis, credit risk, pricing, fraud, or portfolio optimization preferred.
* Strong working knowledge of machine learning algorithms, AWS and cloud development tools, and Data Science/Analytics/BI methodologies.
* Advanced skills in SQL, Python (R or SAS a plus); strong skills in exploratory data analysis, statistical model development, validation/testing, and operation (including performance monitoring).
* Experience with Git/source control (branching, pull requests, code reviews) preferred.
* Advanced knowledge of statistics and experiment design.
* Ability to design, execute, and interpret controlled experiments with…