September 2026 Summaries

Coralogix has introduced a generally available APM experience designed to monitor every service automatically from its first appearance and help users identify whether performance problems stem from releases, environments, transactions, dependencies, or database queries. The platform applies predefined health policies for latency, error rate, cases, and log errors, surfaces breached thresholds directly from alerts, and correlates deployment and custom-event markers with service metrics to speed investigations. Its service catalog, ownership, dependency map, saved views, and signal-status indicators provide teams with centralized visibility into service health, responsibility, and observability coverage. The same APM data can be accessed by Coralogix’s Olly AI assistant, the Coralogix CLI, and MCP-connected IDE agents, enabling automated checks and contextual troubleshooting. The experience requires no new SDK, code changes, or database agents, but depends on compact span metrics from both services and databases, with eligible OpenTelemetry Span Metrics integrations already supporting the capability.

Coralogix describes how its CX CLI enables coding agents to turn Real User Monitoring data, stored as cx_rum log events, into product analytics reports through natural-language prompts. After installing and initializing the CLI, users can request self-contained HTML reports or terminal-friendly Markdown summaries containing traffic KPIs, page rankings, click activity, checkout conversion counts, error tables, and weekly change assessments, without manually writing DataPrime queries or using the UI. The article provides example CLI queries for top pages, clicks, page-level user counts, and errors, while noting that its page-count approach is not a strict sequential funnel and that the Product Analytics Funnels view should be used for that purpose. It also introduces Olly, Coralogix’s AI analyst, for investigating causes behind metrics changes, retrieving generated charts and tables, and conducting follow-up analysis through terminal commands.