For years, the standard approach to robotic process automation was user interface (UI) automation. Without an application integration or application programming interface (API) access, RPA robots would log in to the clients’ business applications just like any other end user and replicate the steps a human took to perform the work. As the business world demanded greater connectivity across their everyday applications, providers made APIs more readily available. Unlike an integration where one application depends on another, an API acts as an interface between two or more existing applications, allowing a developer to define how they communicate with each other.
Until recently, on premise loan origination system (LOS) instances were standard in the mortgage lending industry. With these local instances, every step in a process that a human takes within the application has to be documented and executed within the automated script. Examples of the steps include navigating to a URL, selecting a menu item, typing in a form field, or clicking a button. Because of this, these types of scripts can be bulkier, more susceptible to breakage, and typically have a longer runtime. Naturally, clients expect robots to process loans faster than their human counterparts. Without an API integration between the RPA software and the LOS, speed was one obstacle that was difficult to overcome.
That all changed when Ellie Mae®, creator of Encompass®, the leading loan origination system in the United States, opened API access to select partners. HPA passed their integration partner certification in February 2020, and is currently the only RPA provider in the Ellie Mae partner ecosystem. The feature-rich API to Encompass allows for direct interaction with raw loan data and various Encompass markers and triggers. Accessing this information via API allows RPA robots to bypass the UI, when possible, to dramatically simplify the automated script and process loans faster.
Features and Benefits of the Encompass API
By interacting directly with the data stored on Ellie Mae servers, navigation between various forms and dialogs, and the associated cycle time, is eliminated. Bots are also able to process multiple loans simultaneously on the same machine. In addition to the substantial increase in processing speed, API access reduces process variability by eliminating the many logical branches that are required with UI-based automation. The developer can simply utilize the specific data needed to process the loan to specification and exclude the rest. The resulting automated scripts are more accurate, execute more quickly, and have greater recover-ability in the event of process irregularity.
The utilization of raw data via API also eliminates the user steps associated with “stare and compare” data validation. Combined with field type validation built into the API, multiple checkpoints ensure data consistency throughout the entirety of the loan lifecycle. In the event of an abnormality or error, API technology enables developers to access error messages to determine why the problem occurred. They are then able to build automated catches into the code base so failed processes can try again in a few minutes. This development can be expanded upon if certain processes are consistently problematic. This built-in failure detection helps minimize the burden on human processors while also providing necessary information about the issue for a speedy resolution.
While API access facilitates greater efficiency in automation, some originations processes still cannot be completed 100% with API automation alone. For example, generating and sending disclosures still partially relies on the Encompass UI because these capabilities are not available within the API yet. Thankfully, time savings can still be achieved as both the UI and API can be utilized within automation to execute this small group of processes.
In the fast-paced, highly-competitive world of mortgage lending, every minute shaved off a loan lifecycle matters. Automation already offers lenders 24/7 loan processing, and currently, our bots run pipeline queries every five minutes for loans that meet the criteria for processing. Leveraging our API access to interface with custom webhooks further expands upon this capability, enabling real-time notification that a loan is ready for processing. This allows our team to reduce idle time, delivering lending decisions to borrowers faster than ever.
Today, HPA is increasing throughput with fewer machines, leading to even greater cost savings for our clients. Smaller processes, like fraud detection, can be completed in the time it would take a processor to log into Encompass, navigate to the pipeline, and open a loan from their queue. For more complex processes, average automation runtime has decreased by up to 75%. We look forward to our continued partnership with Ellie Mae, helping our clients achieve greater operational savings and efficiency, and most importantly, deliver a superior experience to their borrowers.