Union Data Scientist Jobs (NOW HIRING)

The Data Scientist supports Credit Union of Texas's vision to be the trusted financial partner for our members and our community by applying statistical, machine learning, and data engineering techniques to credit union data. Working within the Finance function and partnering broadly across the organization, the role develops dashboards, reports, predictive models, and analytics that inform financial forecasting, member experience, marketing effectiveness, and operational decisioning. The Data Scientist collects and integrates operational and market data, identifies trends and variances against forecast, designs and validates models, and operationalizes analytics in partnership with business stakeholders. The role uses CUTX-approved AI and machine learning tools under defined governance, with mandatory human‑in‑the‑loop review on any model output that influences member or financial outcomes.

• Partner with stakeholders across the organization to identify opportunities to leverage CUTX data to drive business solutions and improve member outcomes.
• Mine and analyze data from CUTX databases and external sources to drive optimization of product development, marketing techniques, and business strategies.
• Develop custom data models, algorithms, and feature sets applied to credit union data, including loan, deposit, transaction, and member‑behavior data.
• Apply predictive modeling to support member experience, revenue generation, targeting, retention, and other business outcomes.
• Apply statistical methods (regression, distribution analysis, hypothesis testing) and machine learning techniques (clustering, decision trees, random forest, boosting, neural networks, text mining) appropriate to the business question.
• Develop and maintain financial forecasts, dashboards, and reports that compare actual results against forecast and identify material trends and variances.
• Collect and integrate operational and market data to support financial analytics for the CIO and Finance leadership.
• Translate analytical findings into clear written and verbal narratives suitable for executive and cross‑functional audiences.

• Work with and contribute to data architectures that support repeatable, auditable analytics.
• Assess the effectiveness and accuracy of new data sources and data‑gathering techniques before they are adopted for production use.
• Build and maintain data pipelines, queries, and reusable analytic assets using SQL, Python, R, or other CUTX‑approved tools.
• Develop and operate an A/B testing framework and test model quality before and after deployment.
• Coordinate with functional teams to implement models in production and monitor outcomes against expected performance.
• Develop processes and tools to monitor model performance, data quality, drift, and accuracy over time, and document validation results.
• Document model purpose, data inputs, assumptions, limitations, validation results, and intended use consistent with CUTX model risk and AI governance expectations.
• Coordinate across Finance, Marketing, Lending, Risk, IT, and Compliance to align analytic work with enterprise priorities.
• Continuously develop technical and domain skills, including credit union and financial services knowledge, to improve the relevance and quality of analytic work.

Outcome

Primary KPI

Target / Direction

Analytic work delivers usable, decision‑grade insight to business stakeholders.

Stakeholder‑Accepted Deliverables — percent of analytic deliverables accepted by the requesting stakeholder without material rework.

Production models perform within validated tolerances.

Model Performance Within Tolerance — percent of monitored models operating within defined performance and drift thresholds.

Monthly

Forecast accuracy supports reliable financial planning.

Forecast Variance — absolute variance between forecast and actual on tracked financial metrics.

Monthly

▼ [Target — confirm with CIO]

Models and analytics are fully documented and audit‑ready.

Model Documentation Completeness Rate — percent of in‑use models with complete, current documentation per CUTX standards.

AI‑and ML‑driven outputs influencing business or member decisions are reviewed before use.

Human‑in‑the‑Loop Review Rate on Member‑or Financial‑Impacting Model Outputs — percent of such outputs validated by the Data Scientist (or designated reviewer) prior to action.

Monthly

• Master's degree in Engineering, Computer Science, Data Science, Statistics, Mathematics, Finance, or another quantitative field required.
• Knowledge of the financial services industry is preferred.

• Two (2) to five (5) years of experience manipulating data sets and building statistical and machine learning models.
• Hands‑on experience querying databases and using statistical computing languages such as R, Python, and SQL.
• Experience applying statistical and data mining techniques including GLM/regression, random forest, boosting, decision trees, text mining, and social network analysis.
• Experience designing and using machine learning algorithms (regression, simulation, scenario analysis, clustering, decision trees, neural networks).
• Experience visualizing and presenting data and analytic findings to non‑technical stakeholders.
• Prior credit union, banking, or financial services analytics experience preferred.
• Prior experience with A/B testing frameworks and model monitoring in production preferred.

• No specific license or registration required for this role.

• Strong problem‑solving skills with an emphasis on translating business questions into analytic solutions.
• Proficiency in statistical computing languages (R, Python, SQL) for manipulating data and drawing insights from large data sets.
• Knowledge of data architecture, data pipelines, and data quality practices.
• Working knowledge of a variety of machine learning techniques (clustering, decision trees, random forest, boosting, neural networks, text mining) and a clear understanding of their real‑world advantages and limitations.
• Working knowledge of advanced statistical techniques (regression, distributions, hypothesis testing) and appropriate application.
• Excellent written and verbal communication skills for coordinating across technical and business teams.
• Drive to learn and adopt new technologies, techniques, and CUTX‑approved AI tools.
• Familiarity with model risk management and AI governance concepts is preferred.

Competency

Proficiency Level

Why This Matters in This Role

AI Literacy

Advanced

The role builds, evaluates, and operationalizes AI/ML models (Tier 2) and is expected to recognize when a use case escalates to Tier 3, apply appropriate controls, and ensure human‑in‑the‑loop review.

Analytical Rigor

Advanced

The role is responsible for the statistical and methodological soundness of models that inform financial and member‑impacting decisions.

Advanced

Model error, bias, or misuse can produce direct member harm, fair lending exposure, or financial loss; the role must identify and elevate model risk.

Intermediate

Models influence member experience, targeting, and outcomes; design choices must reflect fair treatment and CUTX member values.

Communication

Advanced

The role must translate complex statistical and ML concepts into clear, decision‑ready narratives for executives and cross‑functional partners.

Operational Discipline

Intermediate

Models, code, and documentation must be reproducible, version‑controlled, and audit‑ready to meet model risk and AI governance expectations.

Compliance Orientation

Intermediate

Analytic work must respect Fair Lending, UDAAP, GLBA data protection, and TRAIGA‑aligned AI governance constraints.

The following workflows are AI‑augmented in this role. The Data Scientist is expected to work fluently within these workflows, exercise sound judgment over AI outputs, and follow all applicable controls.

• Predictive and statistical model development for financial forecasting, member behavior, and product performance.
• Machine learning model design, training, validation, and monitoring on CUTX data.
• AI‑assisted code generation, query authoring, and analytic prototyping using CUTX‑approved tools.
• Automated data quality checks, anomaly detection, and model drift monitoring.
• AI‑assisted summarization of analytic findings and drafting of stakeholder‑facing narratives.

This role operates in AI Tier 2 for its principal AI‑augmented workflows (see Appendix A). The Data Scientist retains accountability for any decision, communication, or member/employee‑impacting action influenced by AI output, consistent with the CUTX Generative AI Usage Policy §3.4.

The Data Scientist is required to:

• Apply human-in-the-loop review on every model output, recommendation, or analytic finding that informs a member-impacting or material financial decision before it is acted upon.
• Validate AI-assisted code, queries, and analytic artifacts before they are used in production or shared with stakeholders.
• Escalate to the CIO and the AI Council any use case that involves direct member-impacting decisioning, automated adverse action, or regulated decisioning, which is treated as Tier 3 and requires additional controls.
• Stop reliance on AI or model output and elevate immediately if the output appears inaccurate, biased, non‑compliant, or outside the role's documented scope (Generative AI Usage Policy §3.5).
• Refrain from entering member non-public personal information (NPI), confidential CUTX information, or material non-public information into any AI tool not explicitly approved for that data classification.
• Complete all required AI training within thirty (30) days of hire and maintain annual currency.

The role is approved to use the following AI tools in performing essential functions (subject to the Generative AI Usage Policy and any tool-specific guidance issued by the AI Council):

• CUTX-approved internal AI assistants (e.g., Sam) for general productivity and approved knowledge tasks.
• Microsoft Copilot for office productivity (drafting, summarization, spreadsheet support).
• CUTX-approved data science and ML platforms and notebook environments (Python, R, SQL toolchains) deployed within CUTX infrastructure.
• CUTX-approved code‑assistance tools used within sanctioned development environments.
• Approved visualization and BI tools used in conjunction with CUTX data.

Use of AI tools outside this list requires prior approval from the role's department leader and the AI Council, per the Generative AI Usage Policy §4.

In addition to the prohibited uses defined in the Generative AI Usage Policy §3.6, the following are specifically prohibited in this role:

• Using model output as the sole basis for any adverse member action (e.g., credit decisioning, account closure, fee assessment) without documented human review and required compliance sign-off.
• Entering member NPI, full account or Social Security numbers, or confidential CUTX data into any AI tool not explicitly approved for that data classification.
• Deploying any model or AI-driven decisioning into production without documented validation, governance review, and required approvals.
• Using unsanctioned third‑party AI services for CUTX analytics, code generation, or member-data processing.

The Data Scientist is responsible for all enterprise compliance obligations applicable to a CUTX team member, including BSA/AML, OFAC, USA PATRIOT Act/CIP/CDD, GLBA and the Safeguards Rule, Fair Lending laws (ECOA/Reg B, Fair Housing Act), UDAAP, Information Security and Acceptable Use, and the CUTX Code of Conduct.

The Data Scientist is responsible for the CUTX Generative AI Usage Policy (TRAIGA / HB 149-aligned), the CUTX AI Playbook (including Tier 2 obligations applicable to this role), and Texas Responsible Artificial Intelligence Governance Act (TRAIGA / HB 149) requirements applicable to the role.

• CUTX Model Risk Management standards covering model development, validation, documentation, monitoring, and change control.
• Fair Lending laws (ECOA/Reg B, Fair Housing Act) as applied to any model or analytic that could influence credit, pricing, or marketing decisions.
• UDAAP standards as applied to member-facing analytics, targeting, and product design.
• Gramm‑Leach‑Bliley Act (GLBA) and the Safeguards Rule governing the handling of member non-public personal information in analytic workflows.
• Fair Credit Reporting Act (FCRA) considerations where analytics interact with consumer report data.
• CUTX Data Governance, Data Classification, and Information Security policies as applied to analytic data sets and code repositories.

This role is performed primarily on‑site at the CUTX Corporate Office, with hybrid eligibility subject to manager approval and CUTX policy. The work environment is office‑based with no significant hazardous or unpleasant conditions. Essential physical activities include the ability to remain stationary at a workstation for extended periods; operate a computer, telephone, and standard office equipment; perform frequent repetitive motions of the hands, wrists, and fingers (keyboard and mouse use); communicate c