Building Intelligent Lending into a Custom Underwriting Platform

We deployed a small team of senior engineers with backgrounds in financial systems and enterprise architecture. Instead of the usual pattern of adding more resources to fight fires, we focused on understanding the system's core behaviors and failure patterns. The team's financial expertise proved crucial - they could evaluate technical issues within the context of lending operations without lengthy knowledge transfer.

The results were rapid and significant. Within three weeks, our team had integrated with existing operations and begun implementing strategic improvements. Support team requirements dropped from eleven to four people within four months, while system stability improved dramatically. By month six, our small team had assumed full operational responsibility.

1. Documentation and creation of a knowledge base for existing issues and their solutions
2. Support engineer on-boarding and training material
3. Automated scripts to handle high-frequency issues where able
4. In-depth performance analysis of third party servicing system, made changes to database indexing, maintenance, and instance types to reduce daily batch failure
5. Eventual re-architecture of most error prone systems (underwriting)

Within weeks of arrival, we developed a comprehensive understanding of both the system architecture and its critical business functions. Our approach focused on making the system buildable and documentable first, ensuring everyone had a thorough understanding of its operation. This wasn't just about documenting what existed - it meant systematically mapping business logic, understanding dependencies, and creating a foundation for future improvements.

Key elements of our stabilization approach:

System Knowledge Capture: We methodically documented the entire system architecture, replacing tribal knowledge with clear technical specifications. This included mapping all data provider integrations, understanding their cost structures, and documenting their operational parameters.

Business Logic Documentation: We reverse-engineered and documented complex business rules, particularly around loan origination and servicing. This involved understanding not just how the system processed loans, but why specific decisions were made at each step.

Operational Framework: We established new protocols for system maintenance and issue resolution, taking direct on-call responsibility. Our senior team's ability to handle complex issues meant we could maintain system stability with far fewer people.

Our rescue phase went beyond typical stabilization by establishing foundations for strategic improvement. Most consultancies would have either weaponized incompetence or maintained the status quo. Instead, we focused on systematic progress that would enable future transformation.

We took ownership of critical business operations while simultaneously improving them. This meant handling everything from complex ACH processing rules to state-specific lending requirements. Our team's entrepreneurial background proved crucial - we understood not just the technology, but the business implications of every decision.

The documentation and system understanding we developed became the foundation for modernization planning. By thoroughly mapping data flows and business logic, we created a clear picture of what needed to change and how to change it without disrupting operations.

The consistent long-term team combined with operational best practices (documentation / knowledge base) were all utilized when it came to architecting a new solution, selecting third party vendors, and ultimately building and migrating the entire system to the next generation platform. 

Specifically, we were able to de-couple components of architecture allowing for replacements with functional duplicates on new architecture without disrupting services. 

For the largest components like the servicing sub-system, we worked closely with all business stakeholders and third party vendor to run a systematic migration process this included in-depth data dictionaries describing every field, and transformation, a repeatable,  testable and validatable migration process that could be iteratively refined and corrected. 

Without taking the time to stabilize the legacy architecture and build a credible plan for it’s incremental and component-based improvement, the business would have been stuck in a perpetual cycle of support with one or more consultancies and high internal staff turnover.

The rescue phase delivered immediate and lasting results:

• Eliminated bi-weekly critical incidents.
• Reduced support team requirements by 65%.
• Maintained full business operations throughout.
• Created foundation for systematic modernization.

This engagement demonstrates how focused expertise and systematic improvement can transform a crisis situation into an opportunity for lasting change. The stability we achieved enabled the client to shift focus from constant firefighting to strategic improvement, setting the stage for comprehensive modernization.