Merge pull request #21997 from newrelic/NR-459576-ai-log-summarization

[feat][Aert AI summary][DO NOT MERGE]: Add AI log summarization enhancements for log-based alerts

Author: adutta-newrelic

src/content/docs/alerts/incident-management/response-intelligence-ai.mdx

@@ -30,7 +30,15 @@ The AI-powered issue page provides a concise summary that includes the affected
   src="/images/nr-ai-summary.webp"
 />
 
-**What’s impacted?**<br/>
+## Access AI summary [#access-ai-summary]
+
+To view the AI summary for any alert issue:
+
+1. Go to **[one.newrelic.com](https://one.newrelic.com) > Alerts > Issues & activity**.
+2. Select an alert issue.
+3. Expand the <DNT>**AI summary**</DNT> section.
+
+**What's impacted?**<br/>
 First responders need to assess the "blast radius" to determine the severity of an issue and decide on the next steps. This widget provides an overview of the affected entities along with the impact to the end users of the application or service.
 
 **What happened previously?**<br/>
@@ -39,7 +47,7 @@ Many IT issues tend to reoccur. Knowing if an issue has happened before, why it
 **What to check?**<br/>
 First responders often need contextual guidance on immediate actions to mitigate an issue. This widget provides customized steps to help them quickly restore services to normal operational levels. Additionally, the Potential causes tab identifies likely causes through causal analysis, covering a range of possible anomalies and performance issues. For more information, refer to casual analysis.
 
-## Causal analysis
+## Causal analysis [#causal-analysis]
 The causal analysis engine identifies potential symptoms that might have triggered an alert event and suggests immediate mitigating actions to address them.
 
 Consider a scenario where a PHP application encounters a memory leak, leading to a failure in the throughput SLI and triggering an alert. Our engine investigates by moving from the service level to the APM application and then to the infrastructure container to detect the symptom.
@@ -55,11 +63,36 @@ At the moment, New Relic only supports causal analysis for APM entities.
   src="/images/potential_issues.webp"
 />
 
-### Mitigating actions & visualizations
+### Mitigating actions & visualizations [#mitigating-actions-visualizations]
 
 For every identified potential cause, the engine offers tailored mitigation actions that guide users through the necessary steps to quickly restore services and entities to their normal operational states. We recognize that many of our customers typically rely on NRQL to analyze significant queries, hence we provide relevant visuals alongside the underlying query for each cause.
 
-### New Relic AI generated analysis
+### New Relic AI generated analysis [#ai-generated-analysis]
 
 In some scenarios, our causal engine may not identify an algorithm-driven cause. However, we have insights that, when combined with with LLMs, can offer you actionable steps. To access this capability, you must have the New Relic AI entitlement enabled.
 
+## AI log summarization for log-based alerts [#ai-log-summarization]
+
+For alert issues triggered by log conditions, the AI summary provides enhanced analysis by automatically examining up to 100,000 log entries from a contextual time window around your alert. This log analysis surfaces patterns, anomalies, and potential causes that aren't immediately visible in the alert itself.
+
+When you expand the <DNT>**AI summary**</DNT> section for a log-based alert, following are details you'll get:
+
+* <DNT>**Analysis context**</DNT>: The number of logs analyzed, time range, and entity name.
+* <DNT>**What happened**</DNT>: A summary explaining the incident with insights beyond the alert itself.
+* <DNT>**Findings (3-4)**</DNT>: Key patterns and anomalies with a specific remediation step.
+
+<img
+  title="The alerts issues feed."
+  alt="A screenshot of the alerts issues
+  feed."
+  src="/images/ai-log-summary.webp"
+/>
+
+To explore findings in detail, click into each finding. This displays comprehensive information below the AI summary section, including:
+
+* Detailed explanation of each finding.
+* Specific remediation steps to resolve the issue.
+
+This log-specific analysis helps you understand potential causes and resolve incidents faster by automatically surfacing critical information from thousands of log entries.
+
+