USU Software AG

07/03/2024 | News release | Distributed by Public on 07/03/2024 02:55

Boosting Efficiency: Pega GenAI Blueprint™ in Action

03.07.2024

Digital transformation in large (and small) organizations is often accompanied by a bouquet of pain points. Big goals and important IT concepts are often slowed down by relatively trivial factors. Many of these are human: fear of change, clinging to traditional methods, fear of employees being "rationalized away" by digital capabilities.

In addition, in traditional, i.e. hierarchically structured companies, cooperation between departments is on shaky ground. The spectrum here ranges from relatively good, occasional exchange to subtlety and communication barriers ("You don't need to know what we do here").

The complex challenge of developing new applications

Developing a new application in this situation is a communicative and organizational risk. Let's take the example of an application for the end-to-end digitalization of a business-critical value-added process.

Four to five specialist departments need to work together, collecting and forwarding data, carrying out checks, giving approvals, and so on. These departments need to find common ground on how to design, build, and run this application. Experience shows that this is no easy task.

Companies that have already made the transition to a process-oriented organization have overcome many hurdles, built new teams, and established agile thinking. They find it much easier to roll out new software across departments because they have the entire process in mind and the goal of delivering the greatest customer value. This unites them, even if they each take on a different role in the process.

But what happens when this is not the case?

GenAI-based tools for improved collaboration in the ideation phase

GenAI-based tools for the design thinking and ideation phase can act as catalysts to significantly improve collaboration. A great example of an AI-based solution is Pega's Blueprint™.

This artificial intelligence tool allows users to quickly create a realistic preview of their planned application. You decide how detailed the description of the application or business process should be. Generative AI makes suggestions at every point, so you can start with purely AI-based descriptions.

Pega's GenAI Blueprint™ is a five-step process that is easy to use, even for novice users. It is available for free on the Web - all you need to do is register with the Pega Community. This makes the tool ideal for a no-obligation evaluation!

Pega Blueprint™ put to the test: A Fictitious Example

We are testing Pega Blueprint™ with a case study for a fictitious insurance company called "Securioblue". The goal is to create a process for the automated processing of auto claims.

We are using the tool in German, although it is originally in English and some translations are still missing. This does not affect our work. Just click on the "Create blueprint" button and off you go!

Step 1: Application context

In the first step, the tool asks for the industry for which the application is needed. We select Insurance.

Click "Next" to proceed to the sub-sector. Here we select "Personal Property and Casualty" and enter "Claims" in the drop-down list below.

Finally, we need to describe what goals or concerns we want to address with the application. In our example, we select "End to End Claims Management" from the drop down list.

GenAI already provides us with a suggested functional description. We can accept this or enter our own text or workflow description. If we had a documented process description, we could use that. This is not the case here.

We decide to formulate our own text because we want to point out to the tool that an expert must always be called for damage to our own vehicle, while this is done for damage to other vehicles depending on the amount of damage.

Click "Next" to start the AI and count the seconds. After 15 seconds, Pega Blueprint shows us a list of case types that match our description.

Step 2: Case Types

In our example, only the first two case types are important. They already contain all the necessary steps. We do not want to have a separate approval or disbursement procedure. By clicking on the corresponding case type, we can check the descriptions generated by the artificial intelligence (AI) and adjust them if necessary.

New case types can be added. Unnecessary task types are deleted.

Even at this stage, an application is generated based on the information provided, which gives all users a very vivid and realistic idea of their future application. In the application, we can easily switch between case types.

All application fields are pre-filled with fictitious data to give a realistic picture. In addition, all views can be viewed separately, be it the desktop, mobile, self-service or contact center view, even the interface view is automatically generated.

Now it's time for the details.

Step 3: Case Life Cycles

Although the AI-generated lifecycle is a good foundation, it is not yet fully developed. Some steps still need to be adjusted.

Clicking on "Edit Life Cycle" brings us to the editing screen. We can move steps back and forth, delete steps or phases, or add new ones.

The case types can also be selected. There are several automated and manual case types available. We save our changes by clicking on "Save". We are satisfied and click "Next".

Step 4: Data model

The AI now generates a list of data fields that are typically required in these case types, based on best practices. We can review and customize these.

Step 5: Define Live Data

The live data is defined in step 5. To do this, a list of data objects is automatically generated that must typically be present in these processes.

One data object is the vehicle itself. This data object already exists in the insurer's backend systems and therefore does not need to be created again. The data object must therefore connect to the insurer's vehicle data system, determine the correct vehicle data, and verify the information provided in the event of a claim.

Step 6: Personas

In this step, we define which roles will interact with the system. In our case, a claims handler, two appraisers, a customer, a payment representative, etc. We will not make any changes here and continue with "Next".

Step 7: Overview

All AI-generated application parts are bundled into one overview page. We have the option to download our blueprint as a PDF or as a blueprint.


If we want to use these building blocks to create a real application for our business, we don't have to start from scratch. We can import the downloaded blueprint into our own Pega instance and seamlessly start the actual application development from there.

Conclusion

GenAI's potential is unique and revolutionary. A key benefit is clear: where previously tedious and lengthy discussions about the correct requirements formulation would have hindered the start of a project, these discussions can now be skipped.

Pega GenAI™ provides a neutral proposal that everyone can accept as a starting point. This dramatically simplifies requirements management because the team no longer works on a blank sheet of paper, but translates any change directly into a realistic application. This makes collaboration much easier.

Another benefit is speed: what used to take hours, days, or even weeks to design or prototype can now be done with a click and a wait of about 15 seconds. Is a department still not satisfied? No problem - just adjust the description, wait 15 seconds while the AI generates new results, and you're done.

As for privacy, let us reassure you: Pega does not and will not use any of the data you enter. The data belongs solely to the users.

Let's talk in person

Want to get your business analysts actively involved in development? Schedule a first meeting with us today!

To date, more than 40,000 blueprints have been created, an impressive testament to the demand and popularity of the tool. Pega GenAI Blueprint™ is based on the extensive knowledge that Pega has gained over the past 40 years in a wide range of industries.

However, it is also clear that Blueprint cannot know your business in detail.

The more specific information you provide to the tool, the more accurate and efficient it can be. From our own experience in creating numerous Blueprints, we can confirm that precise input yields the best results.

One final tip: The tool works for non-Pega organizations as well, as it is available for free on the Web at Pega.com/blueprint. Use the Pega GenAI Blueprint™ as the basis for your next requirements workshop or give it to your IT department. Who knows what miracles it can work?

Share article: