Works with everything you run

The agents speak your stack.

No SDKs, no per-service instrumentation, no replatforming. oneinfra reads from the tools you already deployed. If we don't have a native adapter, the MCP layer lets you wire one up in minutes.

Browse MCP servers →

Observability

Prometheus · Alertmanager · Grafana · OpenTelemetry · Datadog · more

oneinfra reads telemetry directly from your existing stack — no SDKs, no per-service instrumentation.

PrometheusNative

AlertmanagerNative

GrafanaNative

OpenTelemetryNative

DatadogMCP

HoneycombMCP

LightstepMCP

New RelicMCP

Logs

Loki · Elasticsearch · OpenSearch · CloudWatch · Splunk · more

Cluster, application, and audit logs from wherever they already live.

LokiNative

ElasticsearchNative

OpenSearchNative

CloudWatch LogsNative

VectorMCP

MezmoMCP

Sumo LogicMCP

SplunkMCP

Multi-cloud cost

AWS Cost Explorer · GCP Billing · Azure Cost Management · OpenCost

OpenCost-style allocation rolled up with the cloud-billing source of truth.

AWS Cost ExplorerNative

GCP Billing (BigQuery)Native

Azure Cost ManagementNative

OpenCostNative

On-call & incident

PagerDuty · Opsgenie · incident.io · Slack · Microsoft Teams

Send pre-investigated alerts to wherever your on-call already lives.

PagerDutyNative

OpsgenieNative

incident.ioNative

SlackNative

Microsoft TeamsBeta

Code & deploy

GitHub · GitLab · ArgoCD · Flux · Bitbucket

Open the fix as a PR. Tie regressions to commits. Block risky deploys.

GitHubNative

GitLabNative

ArgoCDNative

FluxNative

BitbucketBeta

Kubernetes ecosystem

Pyrra · k8sgpt · eBPF · Cilium · Istio · Linkerd

Plugs into the K8s-native tools your team already runs.

Pyrra (SLOs)Native

k8sgptNative

eBPFNative

CiliumNative

IstioNative

LinkerdNative

Models — bring your own

OpenAI · Anthropic · AWS Bedrock · Azure OpenAI · Ollama · vLLM

Your prompts. Your provider. Your network. Models never see customer data unless you wire them to.

OpenAINative

AnthropicNative

AWS BedrockNative

Azure OpenAINative

Ollama (local)Native

vLLM (local)Native

Don't see your tool?

MCP is the escape hatch.

oneinfra speaks the Model Context Protocol — the open spec for letting AI agents call tools. Any MCP server (yours or off-the-shelf) plugs in via one config line. Most customers wire a custom adapter in under an hour.

See how the agents use them →

Get started in minutes

Put an AI SRE on every cluster.

See oneinfra investigate a live incident, surface idle spend, and draft a fix — in a 20-minute walkthrough on your stack.

Explore the live app →