Pulse Sense (Beta)

Pulse Sense is an AI-powered assistant for Pulse that helps users investigate issues, analyze observability data, and perform operational tasks using natural language.

Instead of manually navigating dashboards, logs, metrics, and alerts, you can ask questions in plain English and receive contextual insights backed by observability data from your environment.

Key Capabilities

• Root Cause Analysis – Investigate application and infrastructure issues by correlating logs, metrics, and workload data.

• Observability Insights – Search and analyze metrics, logs, and operational data using natural language.

• Visualization Support – Create new charts or discover relevant dashboards and visualizations for selected metrics.

• Alert Management – Create, update, and manage alerts using simple natural language requests.

• Operational Actions – Discover and execute predefined operational actions and runbooks.

• Knowledge Discovery – Search operational documentation, runbooks, and knowledge sources for relevant information.

• Anomaly Detection – Identify unusual metric behavior and investigate performance deviations.

How Pulse Sense Works

Pulse Sense interprets your natural language request, retrieves relevant data from the Pulse platform, and generates responses based on real observability data. It can analyze information across metrics, logs, workloads, dashboards, alerts, and actions to provide contextual recommendations and insights.

Example Prompts

Use prompts such as:

• Why are Spark jobs running slowly?

• Why did this Trino query fail?

• Show me failed Airflow DAG runs.

• Create a CPU utilization chart for Trino.

• Alert me when Spark executor memory exceeds 80%.

• Show anomalies in cluster memory utilization.

• Create an alert for these metrics.

• What is causing high resource usage in my cluster?

• List failed workloads from the last 24 hours.