Selecting the Right Decision Engine


Selecting a credit decision engine is a critical decision for any credit grantor and it is not an easy task, as it involves multiple considerations.

Modern decisioning solutions have changed the game for the lending industry. Credit grantors are revolutionising their operations and boosting their competitive advantage by shifting from clunky legacy on-premises decisioning tools to advanced cloud-native decisioning software.

This article is intended to guide lenders in selecting the right decision engine, setting it up for success, avoiding costly implementation mistakes, and ultimately, driving favourable business outcomes.

But how do you find the best decision engine for your business? To help you make the right decision, this article summarises the most important criteria to evaluate when choosing a decision engine.

Please note that for this article the decision engine is assumed to be a Credit Decision Engine.

Pre-Selection Questions and Requirements

It is highly recommended that the credit grantor first makes a list of what their specific requirements will be. This list should be comprehensive and completed prior to engaging with any software vendors:

  • What products and use cases will the decision engine be used for?
  • What payload format does the decision engine need to support?
  • What features will the decision engine need to be configured for?
  • What kind of existing architecture will the decision engine need to fit in with?
  • What kind of data needs to be passed to and from the decision engine?
  • What kind of data access is required in the lender’s use case?
  • Will there be a need for data orchestration?
  • How does the decision engine store and retrieve data?
  • What kind of data retention and archiving policies will the decision engine need to adhere to?
  • How complex will the business rules be?
  • Will there be a requirement for a rules engine?
  • Will there be a need for rule flows?
  • How often will the business rules need to be updated?
  • How technical are the users of the decision engine?
  • What kind of regulatory requirements need to be addressed by the solution?
  • What kind of authentication is required?
  • What kind of data protection and encryption does the decision engine need to support?
  • What kind of infrastructure does the decision engine need to run on?
  • What is the required timeframe for the solution be configured and set up?
  • Do the various connectivity protocols that the decision engine supports match the lenders requirements?
  • What kind of reporting requirements need to be met by the solution?
  • Is there a requirement for the decision engine to have online reports?
  • Is there a need for the decision engine to support batch input/output feeds?
  • What is the uptime SLA required from the decision engine?
  • What are the typical response time thresholds that the decision engine offers?
  • How does the decision engine software vendor perform maintenance?
  • What is the deployment technique that the decision engine uses?
  • How often does the provider roll out upgrades?
  • How does the provider handle major version upgrades?
  • What kind of orchestration requirements are there?
  • Does the lender have custom use cases that will need to be supported by the decision engine?
  • What pricing model does the lender prefer?
  • Does the lender prefer subscription-based or transaction-based pricing?
  • What kind of billing schedule is more preferable for the decision engine to follow?
  • Does the pricing fit within the lender’s budget?
  • What additional fees will there be?

Once this long list of requirements has been completed, then and only then, does it make sense to start the software and vendor selection process.

Selection Considerations:

Integration with Existing Systems 

  • Compatibility
    • Ensure the credit decision engine can integrate seamlessly with your current host systems, such as a Loan Origination System (LOS), Loan Management System (LMS), Cards Processing System, Collections System and customer databases.
    • Verify the decision engine’s compatibility with common file formats and protocols used in your existing systems (e.g., CSV, XML, JSON).
  • APIs and SDKs
    • Look for decision engines that provide comprehensive and well-documented APIs, allowing easy integration and communication between systems.
    • Ensure the SDKs support multiple programming languages (e.g., Python, Java, C#) for flexibility in implementation.


  • Growth Potential
    • Choose a decision engine that can handle an increasing number of credit applications as your business expands, without significant performance degradation.
    • Assess whether the decision engine can process large volumes of data quickly and efficiently, using techniques such as parallel processing or distributed computing.
  • Cloud-Based Solutions
    • Consider cloud-based credit decision engines (e.g., Amazon Web Services, Microsoft Azure, Google Cloud Platform) that offer scalable resources and flexible pricing models.
    • Cloud solutions provide on-demand scalability, ensuring you only pay for what you use, which is ideal for start-ups with fluctuating workloads.

Compliance with Regulations

  • Local Regulations
    • Ensure the decision engine adheres to local financial regulations and standards, such as GDPR (General Data Protection Regulation) in Europe, and any specific requirements for your operating markets.
    • Consider compliance with anti-money laundering (AML) and know your customer (KYC) regulations to avoid legal issues.
  • Regulatory Updates
    • Choose a decision engine that can automatically update and adapt to new regulations and standards, reducing the burden on your compliance team.

User Interface

  • Ease of Use
    • A user-friendly interface is essential for rapid adoption by the credit risk team and other business users. Look for intuitive dashboards, clear navigation, and minimal learning curves.
    • Features such as drag-and-drop rule configuration, visual data representation, and step-by-step wizards can enhance usability.
  • Training and Support
    • Evaluate the availability of consulting and training resources, such as online tutorials, webinars, and documentation, to help your team understand and utilise the decision engine effectively.
    • Ensure that the software vendor provides ongoing support through various channels (e.g., phone, email, chat) to promptly address any technical issues or queries.

Cost Considerations

  • Affordability
    • For any financial institution, and especially for a start-up or small lender, budget constraints are crucial. Look for cost-effective solutions with transparent pricing structures that align with your financial capabilities.
  • Pay-As-You-Go Models
    • Consider decision engines offering flexible pricing models, such as pay-as-you-go or subscription-based pricing, which can help manage costs and avoid large upfront investments.
    • Evaluate the total cost of ownership, including initial set-up fees, licensing, maintenance, and potential upgrade costs.

Flexibility and Customisation

  • Custom Rules
    • The decision engine should enable you to create and customise credit scoring rules and decision criteria to suit your specific market conditions and business model.
    • Features such as rule-based engines or machine learning models can offer greater flexibility in decision-making processes.
  • Adaptability
    • Ensure the decisioning system can adapt to new data sources, changing market conditions, and evolving business needs.
    • Look for decision engines that support continuous improvement and optimisation of credit decision rules based on feedback and performance data.

Data Security

  • Encryption
    • Strong encryption methods for data at rest and in transit are crucial to protect sensitive customer information.
    • Ensure the decision engine complies with all necessary data protection standards.
  • Access Controls
    • Implement robust access control mechanisms to restrict data access to authorised personnel only. This includes role-based access control (RBAC) and multi-factor authentication (MFA).
    • Regularly audit access logs and monitor for any unauthorised access attempts.

Vendor Reputation

  • Track Record
    • Research the software vendor’s experience and success in relevant industries and markets. Look for customer reviews, case studies, and testimonials relevant to your context
    • Consider vendors with a proven track record in your institution’s industry, demonstrating reliability and performance in similar use cases.
  • Local Presence
    • Vendors with a local presence or understanding of your specific market may provide better support and configuration options.
    • Local expertise can help navigate regulatory requirements, cultural nuances, and market-specific challenges.

Technical Support and Maintenance

  • Support Availability
    • Reliable customer support and technical assistance are crucial. Ensure the vendor provides at minimum 8 hours daily support, especially given the time zone differences in many markets.
    • Check the vendor’s service level agreements (SLAs) for response times and resolution guarantees.
  • Maintenance Services
    • Ensure the vendor offers regular maintenance and updates to keep the system running smoothly and securely.
    • Evaluate the availability of patch management and system upgrades to address vulnerabilities and enhance functionality.

Innovation and Future Proofing

  • Machine Learning and AI
    • Consider decision engines that incorporate machine learning and AI capabilities to improve decision accuracy and adapt to new data patterns.
    • AI-driven decision engines can provide advanced analytics, predictive modelling, and automated decision-making processes.
  • Innovation Roadmap
    • Evaluate the vendor’s commitment to innovation and their roadmap for future enhancements. This can include new features, integration capabilities, and technological advancements.
    • Partner with vendors that actively invest in research and development to stay ahead of industry trends and emerging technologies.


By focusing on these detailed considerations, you can select a credit decision engine that meets your needs as a financial services organisation. It is essential to balance cost, functionality, and future growth to ensure a sustainable and efficient decision engine solution.

Examples of Key Considerations in Practice

  • Integration: For instance, if you use a specific Loan Management System, ensure the credit decision engine has native integration capabilities or provides an API that can communicate effectively with the LMS that you use.
  • Scalability: A cloud-based solution such as AWS or Azure offers on-demand scalability, enabling you to adjust resources based on your current needs without significant upfront investments.
  • Cost: Consider decisioning solutions which offer affordable, scalable options with robust support and features tailored for a wide range of credit grantors, products and markets.
About the Author

Stephen John Leonard is the founder of ADEPT Decisions and has held a wide range of roles in the banking and credit risk industry since 1985.

About ADEPT Decisions

We disrupt the status quo in the lending industry by providing clients with customer decisioning, credit risk consulting and advanced analytics to level the playing field, promote financial inclusion and support a new generation of financial products.