Design Partner programme: the first 100 teams get SreNix free. 100 places left.

See the offer and apply

Detect. Diagnose. Remedy. Verify.

Everyone else sends alerts. SreNix sends fixes.

An AI reliability engineer for Kubernetes, AWS, GCP, Azure, and the edge. It fixes problems within the guardrails you set, records every action, and runs on your own AI model — so you stay in control.

Up and running with a single Helm installation command.

One operator, every cloudNo telemetry exfiltrationHelm install, 5 minutes
kubectl get driftreportsCLUSTER-SCOPED CRD
NAME SEVERITY SOURCE SUBJECT
drift-9c1e04f2a77b3d18 warning StuckCertificateRequests CertificateRequest/kube-system/api-tls
# fixer ran, re-diagnosed, closed
DriftReport StuckCertReq — fixer ran OK
DriftReport TLSSecretMismatch — fixer ran OK
DriftReport SecurityDrift — DigestPin PR
  attestation: Ed25519 · auto-merge gate: 5/5 ✓
Re-verify in 60s ...
closed-loop is the default, not a roadmap milestone
One operator. Every cloud. Every stack you already run.
Cloud
AWS · GCP · Azure
K8s distros
Kubernetes · EKS · GKE · AKS · k3s · OpenShift · RKE2
Observability & ticketing
Prometheus · Alertmanager · Grafana · Loki (paid) · OTLP (paid) · OpenProject · Jira · ServiceNow · Slack
K8s-native infra
Vault · cert-manager · CNPG · Rook/Ceph · External Secrets · Kong · ArgoCD · Cloudflare
Trigger sources
K8s informers · Alertmanager polling · Webhook (HMAC) · CronJob resync
AI providers
OpenAI · Anthropic · In-cluster vLLM

The autopilot loop, by default

Five steps, re-run every cycle. Closed-loop is the default mode, not a roadmap milestone.

01

Install

One Helm command, on any Kubernetes 1.27 or newer.

02

Detect

Checks for 21 kinds of Kubernetes problems, plus 10 kinds each for AWS, GCP and Azure.

03

Remediate

Five fixers, each with strict limits. Fixes suggested by AI need a person to approve them.

04

Report

Sends findings to Slack, Alertmanager, Jira and more. You can also read them with kubectl.

05

Verify

Checks again after every fix. No “it probably worked”. It confirms.

See how it works

Limited offer · Design Partners

Be one of our first 100 Design Partners. Use SreNix free.

Download the engine at no cost, let it fix real problems on your clusters, and help shape the product. 100 places left.

The difference

Three hours vs. three minutes

One familiar failure, handled two ways. Brand-new incidents are rare. Repeats make up most of the work.

The existing path

~3 hours
  • T+0Alert fires; the page goes out
  • T+11mA human reaches a keyboard
  • T+45mContext rebuilt from dashboards and chat
  • T+2hFix applied by hand
  • Later“Confirmed” and logged by memory, not measurement

The SreNix path

~3 minutes
  • T+0Probe detects the failure
  • T+30sCause found; checked against your approved list
  • T+1mPolicy-bounded fixer applies the fix
  • T+3mThe same check runs again and confirms it is fixed
  • AlwaysSigned record saved. Nobody was paged.

Why SreNix is different

Four design choices that other tools would have to rebuild themselves to copy.

01

An agent that actually fixes things

Most tools only summarise the problem. SreNix picks a fix, signs it and applies it. Your rules limit what it may do and where.

02

Runs in your cluster, on your own AI model

No cloud service and no lock-in. Use any OpenAI-compatible model. Your data never leaves your network.

03

You can read the code, and every action is recorded

The free Community engine is source-available. Every AI action is signed and can be replayed later, even in the free tier.

04

Evidence pack for auditors

One command builds a pack for your auditors: a list of files, a word-for-word log of every action, and every outcome. It never leaves your network.

vs. the competition

Detailed comparisons

SreNix detects, fixes and checks by default. Other tools help the on-call person you already have.

How SreNix compares with other tools
ProductWhere it runsFixes problems itself?Pricing
SreNixIn-cluster operatorYes, by defaultFlat per-cluster
NeuBirdCloud service that pulls your dataNo (read-only by design)$15–25 / investigation
Resolve AICloud service plus a small agentOn its roadmapContact sales
CiroosCloud service, queries data in placeUnclear (“autonomy slider”)Contact sales
OpenSRE (Tracer)You host it (docker-compose)No (blocks changes)Free, open source

Frequently asked

How is Srenix different from a copilot-style SRE assistant?

Copilots help you investigate, then a person has to do the fix. SreNix fixes the problem itself, inside rules you set. Then it repeats the check that found the problem, to confirm it is gone.

What happens when it makes things worse?

It can’t improvise. It only does fixes from a list you approved, and it never touches secrets. If a type of fix ever gets worse, a safety switch sends it back to manual approval.

Does any of our data leave the cluster?

No. SreNix runs inside your network and sends no data out. If you want the optional AI features, use your own model: OpenAI, Anthropic, or one you host.

Is it open source?

No, it is source-available. Anyone can read the code, including before it touches your cluster. It is free for non-commercial use. Commercial use needs a licence.

Are you SOC 2 certified?

Not yet. What exists today is the evidence trail: every action is signed and can be exported with one command. Certification is the next step.

What does it cost?

A flat price per cluster. No fee per incident and no surprises. See pricing.

Can we put our own brand on it?

Yes. SreNix is built to be white-labelled. See MSPs & aggregators.

On-call should be quieter every week

Helm install in 5 minutes. No telemetry exfiltration. No per-investigation surprises.