FeaturesRoadmap
Join the beta
OPEN BETA · KUBERNETES HEALTH

Whatifyourprodishidingsomething?

Horizon continuously measures the technical debt accumulating in your applications before it becomes a production disaster.

prod-eks / acme-corpLIVE
Tech Debt Score
0 / 100
Manual operations62
Resource efficiency81
Deprecated APIs90
Bad practices58
● 2 critical● 5 warnings
Actionable recommendations
CRIT.
api-gateway — CPU 8× over-provisioned
2000m → 300m
+12 pts
CRIT.
worker — manual kubectl detected
3 out-of-pipeline changes
+8 pts
WARN.
frontend — :latest image
Pin to a SHA256 digest
+4 pts
WARN.
api-gateway — HPA missing
Add HPA (min 2, max 8)
+4 pts
If fixed: 74 → 98 / 100

2-minute install · Read-only · First findings in < 5 min

Free audit

We'll audit your cluster for free.

Leave your contact info. We'll analyze your Kubernetes cluster and deliver a full technical debt report — no commitment required.

No spam. We'll only contact you to schedule your audit.

Get a call to schedule your audit optional

We'll call you to pick a time slot and run the cluster audit together.

What interests you most?

The problem

What you don't see today explodes in production tomorrow.

No objective baseline

Impossible to evaluate infra health without a baseline. You're trusting your gut.

Debt accumulates silently

Every sprint adds invisible debt. It always explodes at the worst time.

Invisible value

Impossible to show what a DevOps engagement delivers over time. No numbers, no proof.

The solution

4 axes. Zero blind spots.

The Tech Debt Score measures what actually matters.

Manual operations

Every manual kubectl is a ticking time bomb. We detect all of them.

💰

Resource efficiency

Your cluster is paying for resources it doesn't use. We show you which ones.

⚠️

Deprecated APIs

The next Kubernetes upgrade will break things. We tell you which ones.

🛡️

Bad practices

:latest images, privileged containers, exposed secrets. We find them all.

In action

Detect. Understand. Fix.

Horizon doesn't just find problems. It tells you what to do and creates the ticket.

🔍

Détection automatique

Scanne chaque workload en continu. Aucune configuration, aucun agent à maintenir.

📋

Diff de manifest généré

Exactement quoi changer, ligne par ligne, avec l'impact estimé en €/mois.

🎫

Ticket créé dans la foulée

Linear ou GitHub Issues. Assigné, labellisé, fermé automatiquement quand le fix est mergé.

📈

Score mis à jour en direct

Chaque correction remonte le score. L'amélioration est visible par toute l'équipe.

Who it's for

Built for those who run production.

Que tu sois CTO qui arbitre les roadmaps, ingénieur DevOps qui gère les incidents ou développeur qui déploie en prod — Horizon te donne la même chose : une mesure objective, pas une intuition.

CTO
My infra is at 87/100. Measured, tracked, defensible. I know exactly where to invest next sprint.
DevOps
I fixed 4 critical findings this morning, score went from 61 to 78. I know exactly where I stand.
Dev
Bad practices get caught before they hit prod. I merge with confidence.
How it works

In production in 2 minutes.

01

Install in 2 min

A single Helm chart. Read-only controller, no write access to your cluster.

02

First findings in < 5 min

Score calculated, recommendations generated. Zero configuration.

03

Tickets created automatically

Linear or GitHub Issues. Closed when the issue is resolved.

What's coming

Horizon today. And what comes next.

The product is live. This is just the beginning.

Available now

Score & Findings

See what your prod is hiding.

Continuous health score on 4 axes
Actionable recommendations
Auto-tickets for Linear & GitHub
End of 2026

Native AI

Your copilot that understands your infra.

Claude & Copilot connected to your findings
Corrective PRs generated automatically
2027

Enterprise

For those who deploy at scale.

On-premise & air-gapped
Custom rules & multi-cluster view

Frequently asked questions

Horizon — What if your prod is hiding something?