Hello. I’m Mario Maldari, Technical Director here at Jama Software. Today, I’ll be walking through a few of our AI features, including our new Relationship Discovery functionality.

One of the biggest challenges our customers face isn’t creating requirements. It’s maintaining complete traceability as projects grow. As products evolve, requirements change, teams change, and knowledge becomes fragmented. Missing relationships often aren’t discovered until a design review, impact analysis, or audit, when they’re much more expensive to fix.

Relationship Discovery addresses that problem by using AI to recommend likely relationships between Jama Connect items, helping teams identify traceability gaps earlier.

Let’s take a look at how this works in Jama Connect.

This is the dashboard view in Jama Connect that shows a straightforward Traceability Information Model (TIM). The TIM defines the expected traceability structure for your project. It is the governing context layer. You can see the flow from high-level stakeholder needs all the way down to tests and defects.

Now we can switch over to the Explorer view, where we have all of our requirements organized by set. Let’s take a look at the stakeholder need requirements, which, by definition in the TIM, should have system requirements associated with them.

If we look at the Trace View, we can see that only a handful of system requirements are actually linked. Any requirements without the expected downstream relationships are automatically flagged by the system.

This level of governance and control allows me to know exactly which requirements and item types need the most review and attention. Now we can actively address those gaps by using the Relationship Discovery feature to find potential related items to link.

In the Document View, I can select a particular requirement, and you’ll see the Jama Connect Advisor panel appear on the right.

There are two options here.

The first is Batch Analyze, which performs a quality analysis on the requirement using INCOSE and EARS standards and returns a quality score. At that point, I can choose to automatically rewrite the requirement by clicking Refine, or I can update it manually. This allows you to ensure you have a high-quality requirement before you begin linking downstream items.

Now let’s choose the Recommended Items to Relate option.

We can select from different item types within the project. In this case, we’ll choose System Requirements, since the project’s Traceability Information Model defines that stakeholder needs should link to downstream system requirements.

Once we initiate the search, a list of recommended items is returned. Each recommendation includes a confidence rating generated by Jama Connect Advisor.

As you can see, the system has returned several suggested requirements to review.

The first few are strong recommendations, while others have lower confidence. Those may still be valid but require human review. Each recommendation includes a direct link so I can quickly examine the suggested requirement and decide whether it should be related.

This keeps a human in the loop while dramatically reducing the time required to identify potential relationships.

It’s especially valuable for complex projects or projects that have evolved over long periods of time. These recommendations are available to all users and remain available for 30 days.

I’m going to select the first three strong recommendations and relate them.

Now you can see they’ve been added as downstream items to the requirement.

This allows me to build a process where I review AI-generated recommendations and create relationships as appropriate while always keeping a human in control.

Jama Connect Advisor also allows users to quickly generate test cases directly from requirements. Let’s see how that works.

I’m going to select a hardware requirement and choose the Test Case Generation option from the Jama Connect Advisor panel.

I’ll select the test context.

In this case, I’ll choose a test type and a test vertical. I’ll also choose to generate test cases with detailed test steps so I have additional information to review.

I can even provide additional context, such as constraints, boundaries, or other prompting instructions.

Now I’ll generate the test cases.

Within a few seconds, I’m presented with nine proposed test cases, each including detailed test steps.

As the human in the loop, I can review each test case for applicability.

If none of the generated test cases meet my needs, I can regenerate them immediately. If I like a test case but want different test steps, I can regenerate only the steps. Or I can edit and expand the generated steps manually.

For example, this Verify Propulsion System Thrust Output test looks like a good fit.

I’ll select where I want the test case created.

Of course, Jama Connect recommends the best location based on my Traceability Information Model.

I’ll accept that recommendation and create the selected test case.

Within a few seconds, I receive confirmation that the test case has been created.

Now I’ll return to my hardware requirement, refresh the traceability view, and you can see that the new test case has been created in the location I selected and automatically linked back to the originating requirement.

This functionality dramatically speeds up the test case generation process by creating context-aware test cases while always keeping a human in the loop.

What previously required hours or even days can now be completed in minutes.

In summary, Jama Connect Advisor enables teams to:

• Create higher-quality requirements with Quality Advisor
• Quickly discover related items with Relationship Discovery
• Rapidly generate traceable test cases

All while keeping humans in the loop.

Together, these capabilities help accelerate product development, improve product quality, and reduce time to market.

Thank you for watching this Features in Five demonstration of Jama Connect Advisor and the new Relationship Discovery feature.

To learn more, please contact your Customer Success Manager or visit jamasoftware.com.

Thank you.