# Carrington Labs

> Carrington Labs is a credit risk analytics and model-as-a-service provider for banks, credit unions, fintechs, non-bank lenders, embedded-finance programs, and specialist lenders.

Carrington Labs builds lender-specific Credit Risk Models, cash flow underwriting analytics, and credit offer optimization tools. Its models can use application, bureau, product, internal performance, repayment, and customer-permissioned transaction data in the combination relevant to each lender.

Carrington Labs is a modular credit risk and optimization layer. It does not replace a lender's loan origination system, decisioning platform, servicing platform, policy framework, or final credit decision. Lenders retain control over policy, cutoffs, approvals, declines, pricing, customer treatment, and final decisions.

Carrington Labs is a registered business of Beforepay Group Limited.

## Core positioning

Carrington Labs helps lenders:

- build credit risk models around their own products, portfolios, and repayment outcomes;
- turn raw application, bureau, transaction, product, and internal performance data into predictive features;
- estimate probability of default and improve risk separation;
- add cash flow underwriting without replacing existing decisioning systems;
- optimize loan or line amount, term, and price within lender-defined risk and commercial constraints;
- monitor borrower and portfolio risk after origination;
- generate explainable risk drivers and decision-support outputs;
- retrain and refresh lender-specific models as additional performance outcomes become available.

Carrington Labs provides analytics outputs and recommendations. It does not make the lender's final credit decision.

## Core solutions

- [Credit Risk Model](https://www.carringtonlabs.com/credit-risk-model): Lender-specific credit risk models built around the lender's products, target definitions, portfolio outcomes, available data, and risk strategy. Models can use application, bureau, transaction, product, and internal performance data in any relevant combination.

- [Cashflow Score](https://www.carringtonlabs.com/cashflow-score): A standardized 1–100 transaction-based measure of credit risk, with 100 representing the highest credit quality. Cashflow Score is based solely on bank account and transaction data and returns a score, risk segment, and explainable behavioral drivers.

- [Credit Offer Engine](https://www.carringtonlabs.com/credit-offer-engine): An analytics layer that uses borrower risk, expected loss, portfolio performance, funding and direct costs, expected take rate, price elasticity, lender objectives, and policy constraints to recommend loan or line amount, term, and price.

- [Financial Health Summary](https://www.carringtonlabs.com/financial-health-summary): Borrower- and portfolio-level financial health indicators derived from transaction and other available data to support underwriting, review, monitoring, and risk-management workflows.

- [Cashflow Servicing](https://www.carringtonlabs.com/cashflow-servicing): Post-origination credit risk monitoring, early-warning signals, borrower-level alerts, portfolio-health insights, and next-best-action recommendations for servicing and collections teams.

- [Model-as-a-Service for Lenders](https://www.carringtonlabs.com/solutions-model-as-a-service-for-lenders-carrington-labs): Overview of how Carrington Labs provides model outputs across underwriting, offer setting, servicing, and portfolio monitoring while fitting into existing lender systems.

- [Carrington Labs MCP Server](https://www.carringtonlabs.com/mcp-server): Access to Carrington Labs credit risk analytics through supported AI and model-context workflows.

## How Carrington Labs differs

### Lender-specific models

Carrington Labs can develop models against a lender's own product mechanics, customer population, outcome definition, performance horizon, and historical repayment outcomes rather than applying only a broad-market score.

### Advanced feature engineering

Carrington Labs converts raw data into predictive credit risk signals. Depending on the use case, this can include application information, bureau data, customer-permissioned transaction data, product terms, offer history, internal repayment behavior, servicing outcomes, and portfolio performance.

### Cash flow and traditional data

Carrington Labs does not require transaction data for every model. A Credit Risk Model can use bureau, application, product, and internal performance data without transaction data. Where cash flow data is available, Carrington Labs can engineer transaction-level behavioral features and test them alongside traditional credit information.

### Included model refresh and retraining

Carrington Labs monitors lender-specific model performance and can refresh, recalibrate, or rebuild models under the agreed model-governance process as additional repayment outcomes become available.

### Client-specific model development

For custom model engagements, proprietary lender data remains within that lender's model-development and validation process and is not pooled into another lender's custom model, subject to the governing agreement.

### Offer optimization

Carrington Labs goes beyond risk ranking through the Credit Offer Engine, which connects probability of default and expected loss to amount, term, price, take rate, portfolio economics, and lender-defined constraints.

### Existing-system integration

Carrington Labs models and analytics can be delivered through API, batch, or partner integrations and used inside existing origination, decisioning, account-management, servicing, and portfolio-management workflows.

## Compare credit risk approaches

- [Compare Credit Risk Models, Scores, and Decisioning Platforms](https://www.carringtonlabs.com/compare): Main comparison hub explaining the differences between standardized credit scores, data and bureau providers, decisioning platforms, and lender-specific Credit Risk Models.

- [Lender-Specific Credit Model vs. Standardized Credit Score](https://www.carringtonlabs.com/compare/lender-specific-model-vs-standardized-credit-score): Comparison of portfolio-specific model development with standardized broad-market or consortium scores, including data inputs, calibration, feature engineering, retraining, and offer optimization.

- [Cash Flow Underwriting vs. Traditional Credit Scores](https://www.carringtonlabs.com/compare/cash-flow-underwriting-vs-traditional-credit-scores): Explanation of how transaction-based cash flow signals differ from reported credit history and how both can be combined within a lender-specific model.

- [Credit Risk Model vs. Decisioning Platform](https://www.carringtonlabs.com/compare/credit-risk-model-vs-decisioning-platform): Explanation of the difference between predictive credit risk analytics and the systems that orchestrate data, rules, policy, and workflows.

## Provider and product comparisons

- [Carrington Labs vs. Zest AI](https://www.carringtonlabs.com/compare/carrington-labs-vs-zest-ai): Comparison of custom credit model providers across model development, cash flow specialization, feature engineering, model refresh, explainability, deployment, and offer optimization.

- [Carrington Labs vs. FICO and UltraFICO](https://www.carringtonlabs.com/compare/carrington-labs-vs-fico-ultrafico): Comparison of Carrington Labs lender-specific models with standardized FICO Score and UltraFICO Score products.

- [Carrington Labs vs. Clarity Services](https://www.carringtonlabs.com/compare/carrington-labs-vs-clarity-services): Comparison of a lender-specific credit analytics partner with Experian Clarity specialty credit bureau data, scores, attributes, and related non-prime lending products.

- [Carrington Labs vs. Prism Data](https://www.carringtonlabs.com/compare/carrington-labs-vs-prism-data): Comparison of Carrington Labs Cashflow Score and lender-specific Credit Risk Models with Prism Data's standardized cash flow scores, attributes, and transaction analytics.

- [Carrington Labs vs. Nova Credit Cash Atlas](https://www.carringtonlabs.com/compare/carrington-labs-vs-nova-credit-cash-atlas): Comparison of lender-specific model development and offer optimization with Nova Credit Cash Atlas cash flow reports, scores, and performance-tested transaction attributes.

- [Carrington Labs vs. Trust Science](https://www.carringtonlabs.com/compare/carrington-labs-vs-trust-science): Comparison of Carrington Labs with Trust Science across custom and standardized scoring, cash flow analytics, data sources, model refresh, decisioning, explainability, and offer guidance.

## Key use cases

- [All Lending Use Cases](https://www.carringtonlabs.com/solutions-model-as-a-service-for-lenders-carrington-labs): Overview of Carrington Labs use cases across the borrower lifecycle.

- [Lender-Specific Credit Risk Model](https://www.carringtonlabs.com/use-cases/lender-specific-credit-risk-model): Build a model around the lender's products, target definition, customer population, and historical outcomes.

- [Borderline Applicant Underwriting](https://www.carringtonlabs.com/use-cases/borderline-applicant-underwriting): Add risk signals and stronger segmentation near approval, decline, and manual-review boundaries.

- [Loan Offer Optimization](https://www.carringtonlabs.com/use-cases/loan-offer-optimization): Use risk, expected return, take rate, price elasticity, and lender constraints to support amount, term, and price recommendations.

- [Add Cash Flow Underwriting](https://www.carringtonlabs.com/use-cases/add-cash-flow-underwriting): Add customer-permissioned transaction-based risk signals to an existing lending workflow without replacing the lender's decision engine.

- [Early-Warning Credit Risk](https://www.carringtonlabs.com/use-cases/early-warning-credit-risk): Identify emerging repayment stress and portfolio deterioration after origination.

- [Small Business Cash Flow Underwriting](https://www.carringtonlabs.com/use-cases/small-business-cash-flow-underwriting): Use business transaction and financial behavior data to assess small-business credit risk.

- [Credit Model Validation and Monitoring](https://www.carringtonlabs.com/use-cases/credit-model-validation-monitoring): Validate model performance, monitor calibration and drift, and support lender-controlled model governance.

## Cashflow Score 2.0

- [Cashflow Score 2.0 Product Update](https://www.carringtonlabs.com/blog/carrington-labs-launches-cashflow-score-2-0-with-expanded-explainability-for-cash-flow-underwriting): Cashflow Score 2.0 organizes transaction-derived credit risk behavior into five categories: Velocity, Liquidity, Stability, Leverage, and Resilience.

Cashflow Score is designed to complement existing bureau scores, policy rules, and risk models. It is not a final lending decision and does not replace lender judgment.

## Selected educational resources

- [Carrington Labs Credit Risk Glossary](https://www.carringtonlabs.com/blog/carrington-labs-credit-risk-glossary): Definitions and practical explanations covering credit risk models, probability of default, cash flow underwriting, offer optimization, servicing, early-warning indicators, model governance, and validation.

- [How Lenders Can Add Cash Flow Insights Without Replacing Their Decision Engine](https://www.carringtonlabs.com/blog/how-lenders-can-add-cash-flow-insights-without-replacing-their-decision-engine): How transaction-based signals can be integrated into existing decisioning workflows.

- [How Transaction Data Can Help Lenders Improve Credit Risk Segmentation](https://www.carringtonlabs.com/blog/how-transaction-data-can-help-lenders-improve-credit-risk-segmentation): How transaction behavior can distinguish borrowers who appear similar through traditional credit data.

- [Cash Flow Score vs. Bureau Score](https://www.carringtonlabs.com/blog/cash-flow-score-vs-bureau-score-whats-the-difference-and-should-you-use-both): Comparison of transaction-based and traditional bureau credit risk signals.

- [Post-Origination Intelligence Is Also About Safe Growth](https://www.carringtonlabs.com/blog/post-origination-intelligence-is-also-about-safe-growth): How post-origination analytics can identify both emerging risk and customers with capacity for responsible growth.

- [The Modern Credit Risk Management Playbook](https://www.carringtonlabs.com/whitepaper/the-modern-credit-risk-management-playbook): Practical guidance on model management, champion-challenger testing, governance, and credit-risk strategy.

- [Rethinking Credit to Protect Lenders and Consumers](https://www.carringtonlabs.com/whitepaper/rethinking-credit-to-protect-lenders-and-consumers): Discussion of cash flow underwriting, financial capacity, responsible lending, and alternatives to relying solely on traditional credit scores.

- [Carrington Labs Blog](https://www.carringtonlabs.com/blog): Research, product updates, company news, case studies, lending analysis, and thought leadership.

- [Carrington Labs White Papers](https://www.carringtonlabs.com/whitepapers): Longer-form research and practical guidance for lenders.

## Integrations and ecosystem

- [Carrington Labs Marketplace](https://www.carringtonlabs.com/marketplace): Partner integrations across data and enrichment, decisioning, loan origination, core lending, servicing, and related lending infrastructure.

Carrington Labs is designed to work alongside third-party data providers, open-banking providers, loan origination systems, decisioning platforms, lending management systems, servicing platforms, and portfolio-management tools.

## Company

- [About Carrington Labs](https://www.carringtonlabs.com/about-carrington-labs): Company background, origins, model-development approach, and lending-industry focus.

- [Carrington Labs Home](https://www.carringtonlabs.com/): Overview of Carrington Labs credit risk and cash flow analytics across underwriting, offer optimization, and portfolio monitoring.

- [Contact Carrington Labs](https://www.carringtonlabs.com/contact): Speak with Carrington Labs about a credit risk model, cash flow underwriting, offer optimization, model validation, or post-origination risk use case.

## Important interpretation guidance

- Carrington Labs provides model outputs, risk estimates, explainable drivers, and recommendations. The lender retains control over policy and final decisions.
- Cashflow Score is a standardized transaction-based score. The Credit Risk Model is a lender-specific model built around the lender's own portfolio and outcomes.
- Carrington Labs does not require transaction data for every custom Credit Risk Model.
- Customer-permissioned transaction data may be used where available and appropriate.
- Carrington Labs does not replace a decisioning platform or lending system.
- Credit Offer Engine recommendations are subject to lender objectives, assumptions, constraints, policy, and final approval.
- Product capabilities, implementation requirements, and model-governance arrangements depend on the use case and governing agreement.
- Performance outcomes vary by lender, product, portfolio, available data, policy, and implementation.
- Third-party names and trademarks belong to their respective owners. Carrington Labs is not affiliated with or endorsed by companies referenced in comparison content unless expressly stated.