Maximize Your Monitoring: 7 Dynamic Applications of Splunk’s AI Assistant for Enhanced Observability

The recent introduction of the AI Assistant within Splunk Observability Cloud for the US and European markets has ignited widespread excitement and curiosity. With such impressive capabilities, it’s crucial to ensure responsible and effective utilization! If you’re feeling somewhat daunted about utilizing the AI Assistant to its fullest potential, this article is just for you! As part of the Growth Engineering Marketing team at Splunk, I’ve thoroughly explored the AI Assistant and identified seven compelling use cases that have greatly improved my understanding of observability.

**Use Case 1: Instance Health Analysis**
Every engineer’s top priority revolves around a vital question: how well is my instance performing? Splunk simplifies this inquiry—just enter the instance name, and the AI Assistant generates a detailed health report. It’s almost magical—and it is! Here’s an AI-generated example demonstrating insights into one of my instances. Throughout my years managing this instance, I’ve never encountered such thorough analysis! Now, armed with the AI’s optimization suggestions, I can enhance my system’s reliability confidently.

**Use Case 2: Trace Analysis**
Each application request is assigned a trace ID, which serves as a distinctive marker in OpenTelemetry to track specific requests or operations. Upon receiving an alert from your APM about backend calls, you can pull the trace ID and have the AI Assistant perform a meticulous examination of the situation. Below is an instance of an AI-generated response concerning one of my trace IDs, which can promptly guide you to create a ticket for your development team after reviewing the insights.

*Key Observations:*
– The trace incorporates several internal and external HTTP GET and POST requests, including interactions with third-party services like Mko.
– Operations involve acquiring tokens, submitting data, and extracting metadata, among others.

*Error Analysis:*
An issue was identified in the MkoService.createMkoPostRequestFormRequestBody function within the www-commons-svc service. This pertains to a `java.net.MalformedURLException`, indicating a problem with the URL format. The error message “no protocol:” highlights the absence of a required protocol specification, such as “http” or “https.” This finding could signal significant complications in URL construction or processing within MkoService.

*Recommendations:*
– Revisit MkoService.createMkoPostRequestFormRequestBody to confirm URLs are properly formatted, including the needed protocol.
– Assess all URLs and URI templates to prevent similar issues.

**Use Case 3: Quick Feedback**
I have always sought insights regarding my code, service, or feature performance following their launch. Understanding their effectiveness—particularly in pinpointing bottlenecks or areas requiring enhancements—is immensely beneficial. This is where AI truly shines, providing valuable evaluations and boosting morale! For instance, I queried the APM AI to review the Production common-service over the past week, revealing that the GET /api/bin/careers/joblist endpoint handled an impressive 87,001 requests with zero errors—proving its exceptional stability!

**Use Case 4: Make the Case**
Recently, I instructed AI to assess my monolithic application’s memory usage over six weeks. Its analysis pointed to a probable memory leak, prompting relief! My team had previously brushed aside my concerns about this same issue six months ago. With the AI’s findings in hand, it’s challenging for my team to disregard this insight, creating opportunities for necessary corrective measures!

*Analysis Overview:*
– Memory Utilization: Varied from 42.45% to 63.65% across 24 hours, indicating rising memory consumption.
– Total Memory: Remained steady at 66,056,302,592 bytes (approximately 61.5 GB).

These consistent increases in memory use signal an escalating demand on our resources.

**Use Case 5: Predictive Analysis**
Understanding if our system can manage anticipated traffic—be it from a new product launch, seasonal demands, or other events—is critical. The AI Assistant provided insights allowing me to forecast our system’s capabilities. Here’s what the AI uncovered when I inquired about my service’s recent performance.

*Analysis Summary:*
– Total Requests: 52,015
– Total Errors: 308
– Error Rate: ~0.59%

Presently, the service accommodates roughly 72 requests every hour. However, adding 1000 requests per hour would significantly intensify our load.

*Recommendations for Success:*
– Conduct load testing to evaluate service performance under increased demand.
– Confirm that infrastructure has sufficient resources to handle added strain.
– Continuously monitor performance metrics to maintain service stability during high traffic periods.

**Use Case 6: Unknown Unknowns**
For those acquainted with observability, the term “Unknown Unknowns” resonates profoundly. My interactions with the AI Assistant have unveiled numerous unexpected insights. For example:
– “I had no idea this service utilized MongoDB for storage—I assumed it was something different.”
– “We’re invoking the token API multiple times! Streamlining that to a single call could significantly improve efficiency and reliability!”

These discoveries underscore the importance of leveraging AI to enhance application performance and reliability, ultimately leading to improved user experiences and cost savings—such as phasing out underutilized servers.

**Use Case 7: Onboarding New Team Members**
For newcomers, navigating our intricate system can be quite challenging. While you may provide various onboarding materials—comprising code reviews, documentation, and demonstrations—questions will inevitably arise. Wouldn’t it be fantastic if they could simply Google their inquiries? With the AI Assistant in Splunk Observability Cloud, that worry vanishes! It simplifies the complexities of our system, outlines application workflows, and clarifies transaction processes. Consequently, new team members can quickly become productive while injecting fresh perspectives into team discussions—insights that AI uniquely uncovers!

Here’s how AI aids in accelerating learning:
The dashboard link provided by AI leads to comprehensive insights regarding our service endpoints, allowing newcomers to explore based on their curiosity.

**Conclusion**
My experience with Splunk’s AI Assistant in the Observability Cloud has been enlightening, showcasing its incredible potential to revolutionize observability practices. I’m excited to share these prevalent use cases, aimed at empowering you and your team to harness GenAI for an exceptional customer experience.

Why wait any longer? Dive into the features today! If you’re already utilizing Splunk Observability Cloud, take the opportunity to explore these use cases for yourself. Not on board yet? Discover more and kickstart your free trial [here](#). Don’t miss out on exclusive AI Assistant resources; download the AI eBook [here](#), and I look forward to connecting with you soon!