We have worked on the front lines to achieve Straight Through Processing for Loan IQ clients for over a decade.  End-to-End solutions have been attempted by the largest banks, consulting firms and solution providers and to this day there is one definitive truth: There is no one size fits all Straight Through Processing solution.

What does exist are solutions that accelerate and reduce risk along your Straight Through Processing journey. Solutions have found specialized success in various areas along the operational process chain, some more than others. When it comes to connecting to Syndication Servicing Systems (Loan IQ, ACBS, AFS) many solutions have fallen woefully short, which is why we developed Onboarding Studio which is uniquely designed to solve a Straight Through Processing bottleneck into Loan IQ.

The problem statement, “utilize modern technology to extract from source documents and populate into servicing to achieve Straight Through Processing” should be divided into two steps.

Step 1 – Capturing data from documents/credit agreements
Step 2 – Processing the captured data into Servicing (in this case Loan IQ)

Even the most resource rich and developmentally mature banks have routinely underestimated the difficulty of importing data into Loan IQ.

Step 1: Capturing data from documents/credit agreements
To date what we have seen in the market are technology firms who are able to demonstrate the utilization of AI, LLMs, Machine learning to extract data from complex documents into outputs which can be pushed into other solutions. In the context of reading credit agreements, there has been some degree of success but projects fail due to the uncompromising requirement of perfect accuracy and severe material risk of mistakes. When a Credit Agreement is read by OCR, LLM or other the confidence interval is never 100%.  An operations user is required to go back and verify the entire document manually which yields some time savings but not nearly as much expected and in turn significantly weakens the ROI of the investment. This may change, and some banks have versions of this kind of parsing in place, but as of today, a meaningful fully automated solution does not exist.

As a point of information, a large tier 1 bank had invested $20 million to use advanced natural language processing to parse their credit agreements and after years of implementation they abandoned the project, opting for a more traditional approach.

Any extraction method applied to credit agreements, e-mail, etc., yields a dataset. When you are able to extract, for example, 500 perfect data points from the credit agreement, you still need to get this data into your servicing system (ex. Loan IQ or other). Without setting up the APIs into Loan IQ you won’t know what Loan IQ requires. You won’t know if Loan IQ needs different or additional data points and you won’t have the requirements for any existing business rules which often require extra data points. Additionally, you will need to translate this extracted data into a format consumable by Loan IQ. Developing the Loan IQ APIs to achieve automation pathways always takes double or triple the amount of time you would expect.

Onboarding Studio has done the majority of that work upfront and is positioned as a flexible modern broker between data sources and Loan IQ.

Step 2: Taking the captured data and processing it specifically into Loan IQ
Loan IQ has several limitations which todays tech solutions do not address:

1. Loan IQ uses XML and SOAP messages.
   • Not many developers and technologies use these anymore, making them instantly incompatible with Loan IQ.
   • Onboarding Studio as a ‘layer’ provides a REST framework, which developers now use as an industry standard, to connect solutions with LoanIQ.
2. There is no explanatory error handling in Loan IQ.
   • You either put the data in or receive a generic error message.
   • Onboarding Studio supplies detailed simplified error message handling allowing faster resolution.
3. There is no persistence when putting data into LoanIQ.
   • When error handling, management of requests are not dealt with, they fail and require resubmission.
   • Onboarding Studio provides persistence which captures and stores so you can resubmit or enrich/correct, and resubmit.
4. When you submit data to Loan IQ there is no transformation.
   • Garbage in – Garbage Out.
   • Onboarding Studio allows dashboard analysis & data transformation, both automatically (by setting up rules), and by letting users identify and enrich data to be input properly, thus streamlining workflow.
5. APIs to get data into Loan IQ are piecemeal.
   • Loan IQ come with basic bare endpoints without business related functionality. For example, if you create a customer with one endpoint that customer is not usable because it has various levels of dependent attributes such as contacts, remittance instructions, servicing groups, etc. which need to be populated.
   • Onboarding Studio has a series of developed APIs which accomplish creating business functionality (create full customer, full deal, full facility, close a deal, transform a full deal, etc.) Working with the low level APIs provided by LoanIQ takes a considerable amount of work to orchestrate them to achieve the conditional automation of business functions you will be expecting. For example, when you create a customer, you need to create addresses, and you can not use those addresses until a profile are created which have further dependencies.
   • Onboarding Studio checks for all business requirements to allow proper creation of data into Loan IQ.

Onboarding Studio is a capability layer for servicing systems, one which saves time and reduces risk out-of-the-box. In the absence of Onboarding Studio, you will need to build connection point architecture, translate data, and manually address data flows and business workflows.

If you would like to discuss the value proposition of Onboarding Studio further and assess its potential role and value in your future Straight Through Processing journey, contact us.

Wes
[email protected]