NOTE: Demo visuals use either blurred real data or synthetic placeholders to protect customer privacy.

At Abnormal, customer feedback drives product innovation. But with feedback spread across tools like Jira and Gong, it can be difficult for product managers to keep up with what customers need, love, or struggle with.

To solve this, Ellie Kloberdanz, AI Systems Architect, created the Product Feedback Pulse Agent: an automated system that compiles feedback from multiple sources into digestible, data-driven summaries.

Understanding what customers are saying about products is crucial for guiding roadmaps and prioritizing feature development. Yet, feedback at Abnormal comes in through multiple, disconnected channels:

• Jira tickets from support teams and technical escalations.

• Gong transcripts from customer calls and renewal discussions.

• Internal notes from sales and customer success.
  

Manually reviewing this information across tools takes hours, and by the time a PM finishes their review, the next wave of feedback has already arrived. Important insights get buried, and decision-making becomes reactive instead of proactive.

The Product Feedback Pulse Agent, built in Glean, automates this entire process. Instead of hunting through Jira tickets or transcribed calls, PMs can simply run the agent or schedule it to deliver automated reports.

Here’s how it works:

1. Input the product name (for example, Account Takeover Protection).

2. Set a timeframe (daily, weekly, or monthly) to define the feedback window.

3. Optionally add keywords (like Okta integration or abuse detection) to focus the query.

4. The agent searches across Jira and Gong for relevant mentions and feedback.

5. It clusters findings into themes (e.g., recurring bugs, feature requests, or praise).

6. It creates an executive summary with highlights, notable accounts, and direct links to Jira tickets or calls.

7. The report can be delivered via chat or email, automatically on a schedule.

The Product Feedback Pulse Agent uses Abnormal’s product taxonomy to ensure accuracy.

• It first cross-references the user’s input with Abnormal’s official product names to avoid mismatches.

• It then queries Jira and Gong for mentions of that product within the chosen timeframe.

• Using AI clustering, it groups related comments into themes, making it easy to see patterns in customer sentiment.

• Finally, it generates a structured summary with highlights, trends, and relevant links, sent straight to the PM’s inbox.
  

Ellie designed the workflow to run seamlessly in the background, giving every product team a consistent, repeatable pulse on what customers are saying.

The Product Feedback Pulse Agent is transforming how Abnormal’s product managers engage with customer data.

• Time savings: Instead of manually digging through tickets and calls, PMs can scan a single summary in minutes.

• Improved prioritization: Themes and trends highlight which issues are most urgent or frequently mentioned.

• Continuous visibility: Automated scheduling ensures teams never lose sight of customer sentiment week to week.

• Data-driven decisions: Replaces anecdotal discussions with clear, evidence-based insights.
  

By automating this process, PMs spend less time gathering information and more time building solutions customers actually need.

What makes the Product Feedback Pulse Agent so powerful is its ability to replace manual, anecdotal feedback loops with scalable, analytical insight. Every week, it creates a living record of what customers are saying, complete with trends, comparisons, and supporting evidence.

This tool is a great example of how AI innovation can improve everyday workflows. By combining technical creativity with a deep understanding of product management challenges, Ellie built a tool that helps Abnormal make smarter, faster, and more customer-centric decisions.