A data integration, at its most basic level, involves Extracting and Transforming data from one system and Loading it into another (ETL). The data and documents to be integrated will go through each of these steps. For the ETL to happen, the data and documents have to be moved from one station to the next. There are different ways to move data through the ETL process. We'll consider them from the perspective of each of the four integration approaches detailed in the Planning for Data Integration topic in this Help Center.
Integration Approaches: How Data Moves
In this approach, the applicant is required to complete an application in a separate system, resulting in the creation of the required records in the target local system. Review the Using a Double Application Approach topic for more details on this approach.
There is no ETL in the traditional sense, as the applicants handle the movement of data and documents themselves. Instead of transferring data between systems, the double application approach has the applicant move these things into the target local system through a different portal. Of course, this is a heavy burden to place on your applicants, so think carefully before resorting to this approach.
Manual Data Entry
In this approach, the staff at your institution manually key in data and upload documents to local systems for each application. Review the Manual Data Entry Approach topic for more details on this approach.
In this approach, there is no ETL in the traditional sense for data, as it all takes place in the staff member's head. The movement of data is routed through the staff member: looking at the CAS application details and typing those into the local system, transforming them to match the local system requirements along the way. The movement of documents is handled one by one, with the staff member retrieving the required documents one at a time and loading them one at a time to the target system, applying the relevant metadata in the local system throughout.
Extracting, transforming, and loading batches of applicant data and documents is called batch processing. Review the Batch Processing and Automated Integration topic for more details on this approach.
The data and documents are moved between the ETL stations by a staff member at your institution. The staff member will manually trigger the extraction of a batch of data or documents, manually trigger an automated transformation process or manually perform the required transformations (if applicable), and manually load the batch into the target local system.
Automating an integration means automating the ETL and the movement of data and documents between the ETL steps. Review the Batch Processing and Automated Integration topic for more details on this approach.
In this approach, the movement of data and documents between the ETL stations is automated (performed by computers). While an automated integration can take a variety of forms - from processing batches on a schedule to real-time processing of individual records - its fundamental difference from the other integration approaches is the automation of the movement of data and documents between the ETL steps. For the extract step, automation can be accomplished by leveraging Liaison's APIs: the WebAdMIT API and the CAS API. Guidance on transformation and how that step can be automated is available in the Transforming Your Data section of this Help Center. Some local systems have transformation tools built right into their data import modules.
You can find more guidance on loading data in the Loading Your Data section of this Help Center. Most local systems offer tools for automating the upload of data and documents.