With all the buzz around generative and agentic AI, organizations of all sizes are looking for ways to take advantage of it today; hoping to drive efficiencies and gain valuable insights.

For our Business Automation Workflow customers, the good news is that some of those capabilities are already here, and more advanced integrations are on the way.

From a BAW Case perspective, one of those things you can enable today is to surface the IBM Content Assistant for FileNet to your users who have to handle complex documents as part of their cases or process tasks.

IBM Content Assistant can be added to your existing BAW FileNet platform that is used for your cases. Once configured, some of its features will show up right away.

No Assembly Required

The first feature is its ability to automatically summarize documents added to the repository, optionally based on their document class, and have that summary exposed as a system level property that is visible when the user selects the Properties action on a document. This means that users might be able to get key information they require without even having to open and read potentially large documents.

The Document Explorer UI View that we include on both our default and custom Case Details pages, is already configured to display your document in the rich document viewer provided by the FileNet Content Services platform. This viewer, in addition to being able to display a large number of document formats, also provides features such as markup, water marks and dynamic data redaction.

With the addition of the IBM Content Assistant, the viewer also now provides an interface to the interactive Content Assistant chat window where you ask it to tell you things about the document without having to try and find it yourself.

Some Assembly Required

The way the default case UI views display the document viewer is through one of the URL APIs to Navigator described in the Navigator Knowledge Center here.

The API is called bookmark.jsp and to display a document in the viewer configured for the Navigator repository you just need to pass in the Navigator Repository ID (note that a Navigator repository is a concept one level up from an Object Store) and the document’s FileNet ID.

Here is a simple example of that same API on a click event handler on an icon view that gets the document ID from another view that is bound to a document ID variable.

In this example, the the Navigator server shares the same server/port since it is running in the same server context as the Case Client plug-in so a relative URL can be assembled.

var navRepoName = "bawTosRepo"; // Use your ICN Repository ID here - getting it from a project's environment variables is good practice

var docID =  ${docID}.getValue().substring(4); // Remove the "idd_" CMIS prefix if the value came from CMIS APIs

var myURL = "/navigator/bookmark.jsp?"; // Using server relative URL

myURL = myURL + "&repositoryId=" + navRepoName;

myURL = myURL + "&docid=" + docID;

window.open(myURL, '_blank');

When this URL is displayed in a popup window, it will contain the Navigator Viewer. This viewer, when the IBM Content Assistant is configured, will include the Content Assistant chat UI where you can inquire about the document being displayed.

When using this on a Client Side Human Service (CSHS) that is part of a BAW Case project (eg. a case details page or a BAW process task sunning under a case activity), you can get the icnRepoId value from the caseInstance variable returned from the Get Case Instance Object service call. This call is part of the custom case details CSHS you get when you add a custom case details to your case type.

If you need these values in a stand alone BAW process project, you can define values in the project's Environment Variables section so they are not hard coded into any scripts or view configurations.

I hope this gives you some ideas on how you can incorporate this exciting technology into your BAW projects.