Observability vendors all found religion in AI

The homepage became the roadmap

On 22 Aug, the category's front doors changed almost in unison. Datadog moved to 'AI-Powered Observability and Security'. Grafana Labs went with 'AI for observability. Observability for AI. One platform for both.' New Relic changed its hero to 'End visibility silos' and added navigation for 'AI Observability'. PagerDuty led with 'Move fast and MAKE SURE it works' and copy about incident management that gets smarter over time. Sentry tightened to 'Code breaks, fix it faster' while adding 'Agent', 'Autofix' and 'AI Code Review' to navigation.

That timing matters more than any single line of copy. The market did not merely add AI pages. It rewired the first-screen promise across monitoring, incidents, application errors and platform observability in the same window. The new centre of gravity is not dashboards. It is whether software can understand production context, recommend action and, increasingly, participate in the response loop.

The useful signal is the simultaneity. A founder watching only launches would have seen scattered AI announcements. Watching positioning changes shows the competitive move earlier: the category is trying to teach buyers that AI is now part of observability architecture, not an add-on feature.

Grafana made the cleanest product move

Grafana Labs had the most coherent fortnight. On 22 Aug, press coverage reported general availability for gcx and MCP Server for telemetry-driven agent development, with Grafana tools allowing AI agents to query live observability data. The same day, new YouTube material covered Grafana Assistant, MCP and gcx, alongside product education for Tempo and Digital Experience Monitoring and a PointsBet Grafana Cloud case study.

On 23 Aug, Grafana Labs added eight repositories, including alloy-workshop, ai-sdk, alloy-scenarios, grafanacon2026-opentelemetry-instrumentation, agento11y, faro-android-gradle-plugin, faro-react-native-sdk and backstage-plugin. Its top repositories also showed an increase of 648 stars, reaching 755 in the observed set. On 22 Aug, the company listed 146 engineering roles, including Associate Observability Architect and Backend Engineer, Platform, Stacks, across remote locations including the US, Ireland, Spain and Sweden.

This is the strongest pattern in the week because it connects product availability, developer education, repository activity and hiring. Grafana is not just saying AI agents should use observability data. It is putting names, artefacts and workshops around that idea.

Datadog widened distribution while holding the enterprise line

Datadog's most concrete external move was distribution. Vertosoft announced on 18 Aug that it would bring Datadog's cloud monitoring platform to public sector and government clients. That matters because government adoption is a channel problem as much as a product problem, and the observed move was explicitly about making Datadog's monitoring available for that market.

The product message moved in parallel. On 22 Aug, Datadog's homepage shifted to 'AI-Powered Observability and Security', with calls to action including 'GET STARTED FREE', 'Free Trial', 'Get Started' and 'SEE THE PLATFORM'. Its navigation continued to expose a broad product surface, including Infrastructure Monitoring, Metrics and Cloud Cost Management.

The hiring footprint still dwarfs the field in this sample. On 22 Aug, Datadog had 448 active postings: 223 in engineering, including AI Research Engineer and AI Research Scientist roles, 136 in sales and 54 in marketing. The week also brought YouTube content around Terraform with Datadog, customer cases including Cleo Health and AssemblyAI, 'This Month in Datadog' and AI-driven operations. The message is breadth, not a single wedge.

Incidents, code and infrastructure are converging

PagerDuty used the fortnight to make incident response sound more autonomous. On 22 Aug, the site copy emphasised end-to-end incident management that gets smarter over time, highlighted AI agents and added an 'AI at PagerDuty' navigation section. On 24 Aug, it published a blog post titled 'Self-Improving System: 5 Steps'. The company also had 29 job postings on 22 Aug, split across 17 marketing roles, three AI/ML engineering roles and nine sales roles.

There was also market-facing financial attention around PagerDuty. On 23 Aug, multiple articles said PagerDuty would announce financial results after the market close on 27 Aug. Truist raised its price target from $11 to $13 while maintaining a buy rating, and filings or articles noted share purchases or increased holdings by Pacer Advisors, Glenmede Trust Co. NA and Empowered Funds LLC.

Sentry's signal was more developer-native. On 22 Aug, its homepage and navigation pushed faster fixing, Agent, Autofix, AI Code Review and Session Replay. On 23 Aug, Sentry released version 0.47.0 of the warden repository, adding configurable provider base URLs and surfacing verifier rejections across reports and CLI output. It also created repositories including warden, snuba, sentry-dotnet, perfview and se. On 22 Aug, Sentry had 37 active postings, 29 of them engineering roles focused on machine learning and AI.

New Relic sat between these poles. On 22 Aug, it changed its headline to 'End visibility silos', added 'View Platform' and 'AI Observability' navigation items, and introduced headings including 'Observability built to understand AI' and 'Intelligent Observability'. On 23 Aug, it released version 0.21.1 of the newrelic-k8s-metrics-adapter with dependency updates including k8s.io/utils and k8s.io/kube-openapi, and created repositories including newrelic-k8s-metrics-adapter, nri-prometheus and nri-flex. Its 53 active postings on 22 Aug were concentrated in engineering and sales, including enterprise Account Executive roles.