ChatGPT SearchOct 12, 2026
Loading…
Data as of Oct 12, 2026A topic in AI Observability and Incident Management.
Reviewed by Dimitry Apollonsky
Mentioned in · Jul 5 – Oct 12, 2026
ChatGPT SearchOct 12, 2026
Yes. There are now several observability tools that use AI to turn a service's existing telemetry — metrics, logs, and traces — into useful monitoring dashboards, rather than requiring engineers to build every panel and query manually.
The key distinction is whether you want AI to generate a dashboard on demand or automatically create one whenever a new service appears. Both approaches exist, but they aren't the same capability.
1. Dash0 — best fit for telemetry-aware generation
Its Agent0 AI assistant builds dashboards from actual telemetry, generates PromQL queries, chooses visualizations, and validates that queries return data. It can also generate Terraform configuration for version-controlled dashboards.
Best for: teams using OpenTelemetry and Prometheus-style metrics.
2. ClickStack — best for exploring unfamiliar telemetry
Its AI investigates your data, generates and validates queries, and creates editable dashboards. It can even create linked dashboards for service health and deeper investigations.
Best for: teams wanting AI-led discovery, not just chart generation.
3. SigNoz — best for an AI-agent workflow
Through its MCP integration, a compatible AI assistant can check whether a new service is sending telemetry, discover relevant metrics, and create a dashboard with request throughput, error rate, p99 latency, and other useful signals.
Best for: teams that want to automate dashboard creation using an AI agent.
4. Grafana Assistant — best if you already use Grafana
It can use natural-language requests to create dashboards, refine panels, and work with available telemetry and data sources. You can also customize suggested dashboards to fit your environment.
Best for: avoiding an observability-platform migration.
5. Datadog Bits AI — best if you already use Datadog
Bits Chat can generate dashboards and notebooks from natural-language instructions, using relevant metrics, logs, and traces to create visualizations.
Best for: teams that want AI capabilities built into their existing monitoring platform.
If the goal is zero-touch dashboard creation whenever a new service is deployed, I'd evaluate Dash0 and SigNoz first, then compare them with your existing monitoring stack. Their documented capabilities make them promising candidates, but the full deployment-triggered workflow may require additional automation.
A practical implementation would look like this:
1. New service deployed
Service identity and instrumentation are detected
2. Discover available telemetry
Inspect metrics, logs, traces, labels, and available queries
3. AI generates the dashboard
Choose service-specific KPIs, create panels, validate queries
4. Publish and maintain
Save the dashboard, apply team standards, and alert on missing telemetry
The important caveat: AI cannot reliably create meaningful panels for telemetry that isn't being collected. I'd also require query validation, consistent dashboard templates, and a review or policy gate before automatically publishing dashboards in production.
What monitoring stack are you using today?
Current monitoring stack
Prometheus + Grafana
Datadog
New Relic
OpenTelemetry or another stack
Starting from scratch
Desired automation
AI suggests or builds dashboards when requested
A dashboard is created automatically on every deployment
Discover new services and keep dashboards updated
Deployment environment
Kubernetes
Cloud services / serverless
VMs or bare metal
Mixed environment
[Find the best fit]
Google AI ModeOct 12, 2026
Yes, modern observability platforms have introduced AI copilots and generative SRE agents that build or draft complete monitoring dashboards directly from natural-language prompts or active telemetry streams.
The primary enterprise tools offering AI-driven dashboard and widget generation include:
ChatGPT SearchOct 12, 2026
If your goal is to stop manually building dashboards panel by panel, I'd focus on observability platforms that can generate dashboards from natural-language prompts, discover telemetry automatically, or provide service-specific dashboards out of the box.
There are some strong options, but they differ in how much they automate: some use AI to build a dashboard on demand, while others automatically generate useful visualizations once you connect your services.
(Best overall for existing telemetry stacks)
Describe the dashboard you want in plain English, and Assistant can generate panels, write queries, validate that they return data, and help arrange layouts. It supports a broad range of sources, including Prometheus, Loki, Tempo, CloudWatch, Datadog, and New Relic.
Google AI ModeOct 12, 2026
ChatGPT SearchOct 12, 2026
Yes. As of October 2026, several observability platforms use AI to generate dashboards from natural-language instructions and actual telemetry data. A smaller but important distinction is whether the platform can discover a newly instrumented service and build a useful dashboard with minimal human input, rather than simply generate charts after an engineer describes what they want.
The strongest matches I found are Dash0, Datadog, ClickStack (ClickHouse), and Grafana Cloud. They approach the problem differently, from telemetry-aware dashboard generation to AI assistants that create complete dashboards on demand.
Google AI ModeOct 12, 2026
While these tools drastically cut down manual setup, they usually rely on a clear service identifier or naming convention already reporting data to the platform so the AI knows what telemetry streams to target.
If you'd like to explore this further, tell me:
I can give you more specific steps on how to trigger their AI generation features.
Best for: Teams that already use Grafana or want to unify telemetry from multiple vendors.
A strong choice if you want dashboards closely connected to service topology and full-stack observability. Dynatrace offers ready-made dashboards, natural-language exploration through Dynatrace Intelligence, and Davis CoPilot-assisted dashboard creation. It can correlate metrics, logs, traces, and events.
Best for: Large, complex environments where understanding service dependencies matters as much as visualizing metrics.
Particularly useful if you want to minimize setup from the start. New Relic automatically provides visualizations for instrumented applications, including response times, throughput, errors, and tracing insights. New Relic AI also lets you query telemetry and create charts using natural language.
Best for: Teams that want automatic starting dashboards rather than building every service view from scratch.
Worth evaluating if you want an integrated monitoring platform with service-level visibility and prebuilt integrations. Datadog is a good candidate for reducing dashboard maintenance through its existing dashboards and monitoring workflows, though I would verify the exact AI dashboard-generation capabilities available in your plan before choosing it specifically for prompt-driven creation.
Best for: Teams prioritizing integrated infrastructure monitoring, APM, logs, and operational workflows.
| Your priority | My pick |
|---|---|
| Generate dashboards from prompts | Grafana Assistant |
| Automatically get useful views when onboarding services | New Relic |
| Correlate telemetry across complex service dependencies | Dynatrace |
| Use an integrated, managed monitoring platform | Datadog |
One important distinction: AI-generated dashboards aren't the same as continuously maintained dashboards. Most tools still need telemetry collection and data sources configured. You'll also want to validate queries, metric labels, and service-level objectives rather than assuming the AI knows which signals are business-critical.
If you're already collecting metrics with Prometheus and logs with Loki, I'd start with Grafana Assistant. Try a prompt like:
Example prompt for an AI dashboard builder
Create a production health dashboard for every service in my environment. Include request rate, error rate, p50/p95/p99 latency, CPU and memory usage, saturation, dependency failures, and recent deployments. Use existing telemetry sources, create reusable service and environment filters, validate all queries, and flag metrics that aren't available instead of inventing data.
[Copy prompt]
Grafana documents support for natural-language dashboard creation and query validation.
One question that would help me narrow this down: What are you using to collect telemetry today?
1. Current monitoring stack
Prometheus / Grafana
OpenTelemetry collector / backend
Datadog
New Relic or Dynatrace
CloudWatch / Azure Monitor / Google Cloud
Mixed stack or something else
2. Deployment preference
Self-hosted / Kubernetes
Managed SaaS
Either
3. What should AI automate?
Generate dashboards automatically
Discover services and metrics automatically
Update dashboards as services change
All of the above
[Find the best fit]
Modern observability platforms are shifting away from manual widget dragging and query writing, moving instead toward natural language generation and automated context-mapping . While fully "zero-touch" autonomous dashboard generation for every custom niche service is still maturing, several leading tools significantly cut down setup toil using AI.
No matter which platform you choose, AI-driven dashboard creation is only as good as the metadata attached to your incoming telemetry. To maximize success:
http.server.duration or db.system ), AI assistants can map out relationships and generate accurate dashboard widgets much faster.If you want to narrow this down, let me know:
(Strong telemetry-aware match)
Dash0's Agent0 generates dashboards using real telemetry, writes PromQL queries, chooses visualizations, and validates that queries return data before creating panels. It supports service overviews, RED metrics (request rate, errors, duration), and multi-service comparisons. It can also generate Terraform configuration for dashboards.
Best for: Teams using OpenTelemetry and Prometheus-style metrics that want reliable, service-specific dashboards.
(Strong general-purpose match)
Bits Chat can generate complete dashboards from natural-language prompts, using relevant metrics, logs, and traces to assemble charts and queries. Engineers can iterate conversationally to add panels, change queries, and reorganize the dashboard. Datadog also has conventional auto-generated and integration dashboards.
Best for: Organizations already using Datadog that want engineers to go from discovering a service to a working dashboard with little manual query writing.
(Strong investigative workflow match)
Announced in July 2026, ClickStack's AI dashboard generation explores telemetry, generates queries, validates visualizations, and creates editable dashboards. It can also create linked dashboards for system-wide health and service-level drill-downs, retaining the investigation that produced them.
Best for: Teams that want AI to investigate unfamiliar telemetry and produce not just dashboards but a navigable troubleshooting workflow.
Grafana's AI assistant supports natural-language queries and dashboard creation. Separately, Grafana's suggested dashboards feature recommends existing dashboards based on connected data sources. The distinction matters: AI-assisted generation is conversational, while suggested dashboards rely on prebuilt templates relevant to the configured data source.
Best for: Teams invested in Grafana, Prometheus, and a broad mix of telemetry backends that want to preserve flexibility.
Axiom's AI dashboard generator takes a dataset and a plain-language description, then generates a dashboard with relevant charts, fields, and filters. Users can refine the result afterward. Its documented approach is centered on generating dashboards from selected telemetry data, rather than proving that dashboards are automatically created whenever a new service appears.
Best for: Teams looking for fast, AI-assisted exploration of logs, events, and other telemetry.
If your goal is: “A new service starts emitting telemetry, and the platform automatically creates a useful monitoring dashboard without an engineer writing a prompt,” I'd distinguish the products as follows.
| Capability | Strongest candidates |
|---|---|
| Generate dashboards from actual telemetry | Dash0, ClickStack |
| Generate dashboards through natural-language prompts | Datadog, Dash0, ClickStack, Grafana |
| Discover useful visualizations and validate queries | Dash0, ClickStack |
| Recommend existing dashboards from connected data sources | Grafana |
| Create linked dashboards for troubleshooting workflows | ClickStack |
| Generate dashboard infrastructure as code | Dash0 |
These are based on documented capabilities, not independent benchmark results.
One important caveat: None of these cited features alone establishes the complete hands-off workflow of detecting every newly onboarded service, inferring its purpose, creating a dashboard automatically, and maintaining it as telemetry evolves. That specific end-to-end behavior should be tested in a proof of concept.
Give each platform telemetry from a previously unseen service and evaluate whether it can:
For an OpenTelemetry-centric environment, I'd start with Dash0 and ClickStack, then compare Datadog if you're already using its ecosystem. Grafana is particularly worth evaluating if you prefer an open, multi-backend monitoring stack.
One question that would help narrow this down: Do you want dashboards generated automatically as soon as a new service appears, with no prompt or human intervention, or is an AI assistant that generates them on request acceptable? That distinction significantly changes which platform is the best fit.
Yes. The industry is actively shifting from static, manually-crafted visualization boards toward an AI-driven operational reasoning layer . While the ultimate vision—an AI passively spinning up fully customized, comprehensive production dashboards the exact millisecond a new OpenTelemetry (OTel) trace or metric hits the ingestion gateway —is still evolving, major players and emerging tools are closing the gap.
Most enterprise platforms handle new services via auto-discovery (instantly mapping service topologies and routing telemetry to pre-built, out-of-the-box standard templates), while leveraging generative AI assistants to build custom panels on demand via natural language.
Industry consensus (highlighted heavily across SRE and platform engineering spaces) points to a structural shift:
If you are looking to implement this in your stack, would you like to explore OpenTelemetry-native backends or are you evaluating a specific platform like Datadog, Dynatrace, or Grafana?