All TRUE implementations begin with a Discovery and Pre-Kickoff stage, where we explore your requirements, discuss the intended outcomes, and establish your success criteria.  

However, before we dive into the implementation process, let’s take a step back and consider the most fundamental question when purchasing any business software:

How will it deliver a return on investment (ROI)?

Traditional software is usually specific about its outcomes. Value can be measured in direct cost savings, process efficiencies, increased productivity, or new capabilities. For example, a new accounting system may offer improved connectivity to online banking systems, saving time and improving accuracy for an accounting team as manual data entry tasks are eliminated. The benefit is specific, the timesaving and accuracy benefits are relatively easy to predict, and the ROI calculation is straightforward. 

The ultimate outcome in a lending process is also specific – a decision to fund a mortgage loan or not. However, the final determination is a calculation based on all the data received during a loan application and analyzed through the loan manufacturing process. This data is variable and so therefore is the process of arriving at a decision. 

In a conventional process, where data is captured and processed with all or mostly human skill, the objective is to gain sufficient data to approve or refuse a loan.

However, lenders cannot easily assess the effectiveness of their data assembly: how accurately was information entered or transcribed, how thorough were verification processes, was everything carried out according to internal standards and regulatory requirements? 

Lending intelligence delivers the required data and transparency into the process and accuracy scores that reveal its quality. Introducing AI into data generation enables lenders to make decisions to approve or refuse loans according to a degree of confidence that can be precisely determined and audited at any time. 

The point is that lending intelligence is less a specific new capability or process and more an improved means of generating the data that drives multiple capabilities and processes (existing, new, and adapted). Specific outcomes can be identified for lending intelligence, but the outcomes are broader and more variable.  

With most AI implementations, including lending intelligence, the ROI generated will depend on factors such as how you configure the software, the engagement of stakeholders in your team and other affected teams, and your effectiveness in redesigning processes to take advantage of your new capability. 

TRUE is proven to deliver borrower data that is significantly more accurate than human processes. The difference compared with traditional software is that the degree of confidence in that data depends on how you configure the AI. You can expect TRUE to reduce costs and timescales, increase manufacturing volumes, give you a more elastic business model, and more. However, it is you who will choose the system configuration and determine the level of comfort in your organization for data confidence and accuracy. 

A process study gives you a detailed understanding of your existing loan manufacturing process upon which you can envision and design a new process that integrates lending intelligence. This clarity about what your AI project is trying to achieve, backed by data from the study, allows us to move into stage 2 of the implementation project: Analysis and Statement of Work (SOW).   

The SOW will be the taskmaster for the remainder of the project. It defines the objectives and limits of your implementation and sets a pragmatic roadmap for how these outcomes will be delivered. It gives granular detail about your requirements, provides measurement criteria for assessing outcomes, and helps us predict ROI with greater certainty.   

While it’s always helpful to retain some flexibility in any project plan, the SOW provides clear guardrails that help keep a project on track and ensure its intended outcomes are delivered. It’s therefore worth taking a moment to use the knowledge acquired in your process study to revisit the questions raised in the Discovery and Pre-Kickoff stage. Settled opinion on the project's objectives will help avoid later questioning of the SOW, allowing it to proceed efficiently.  

Another preparation that will help you to uphold the integrity of the SOW is assessing who are the stakeholders in your project and how their goals (stated and unstated) align with your project goals. Understanding who is involved, what they care about, and how the introduction of AI might affect them will help you to coordinate support for your project and ensure it has the full backing of senior management.

You’ll gain a better appreciation for the influences and “red lines” of different stakeholders, enabling you to navigate the personal motivations and internal politics that are a factor within any organization.   

Analysis of the study will also help you gain a more practical grasp of the potential and limitations of AI-powered lending intelligence. Lending intelligence is not an end in itself. Rather, it is a resource that has a powerful impact on a broad array of processes and commercial agendas that can be improved, redesigned or even eliminated as a result. It’s helpful and important to understand the implications of AI performance versus costs as you configure your implementation.