1 / 19
Stakater Multi-Tenant Operator

From Kubernetes Clusters
to Enterprise Platform

Turning complexity into control, governance, and scale —
without building it yourself.

Red Hat Certified Operator OpenShift Native AKS Compatible Any Kubernetes Distribution

The Strategy That
Seems Right — But Fails

"Let's give each team their own cluster.
That ensures isolation and flexibility."

Sounds right on day one.
By month six, the truth arrives.

Team A
cluster
Team B
cluster
Team C
cluster
Team D
cluster
Team E
cluster
Team F
cluster
Team G
cluster ⚠
Cluster N…
⚠ ⚠

10 → 50 → 100 clusters. Each needs upgrades, policies, access control, monitoring.

What actually happens

NO STANDARDIZATION
Every cluster is a snowflake. Policies drift. Security baselines diverge.
COSTS EXPLODE
Duplicated tooling, idle nodes, no chargeback. Finance can't explain the bill.
PLATFORM TEAM BURNS OUT
Every new team = manual cluster setup. The bottleneck becomes the platform team itself.

This Is Not a
Future Risk. This Always Happens.

6×
more clusters than planned
within 18 months
40%
of platform team time spent
on cluster maintenance
0
teams with clear cost
visibility per workload

Operational cost grows linearly

Add a team → add a cluster → add a full operations burden. There is no economy of scale with cluster sprawl.

Complexity grows exponentially

Security policies drift. Tooling duplicates. Audit trails fragment. The more clusters, the harder compliance becomes.

Clusters don't scale. Platforms do.

You're Solving
the Wrong Problem

Most teams diagnose the symptom — too many clusters — and prescribe more cluster tooling. That doesn't solve it. It scales the pain.

YOU THINK
"We need better cluster management"
THE REAL NEED
A platform layer on top of Kubernetes

Three truths the industry has learned

🚫
Namespace ≠ multi-tenancy
A namespace is a label. Isolation requires policy, network, RBAC, and quotas enforced together.
🚫
Kubernetes gives primitives, not a platform
RBAC, NetworkPolicy, ResourceQuota — these are building blocks. You still have to assemble the platform.
🚫
Self-service without guardrails is a compliance risk
Freedom without policy enforcement is just controlled chaos with extra steps.
This is not a Kubernetes problem.
It's a platform problem.

There Is a
Better Model

Instead of many clusters…

Cluster
A
Cluster
B
Cluster
C
Cluster
N…
↓

Few clusters. Many secure tenants.

Cluster (1–3)
Central governance · Policy enforcement
Team A
isolated tenant
Team B
isolated tenant
Team C
isolated tenant
Team N
isolated tenant

What the platform model delivers

✓
Isolation without cluster per team
Policy-enforced boundaries — RBAC, network, quotas — per tenant, automatically.
✓
Self-service with guardrails
Teams onboard themselves. Compliance is default behavior, not a checklist.
✓
Central governance, distributed execution
One control plane. Every team works inside defined policy. No drift.
✓
Platform team scales without headcount
Add a team in minutes, not days. Automation absorbs what used to require humans.
Clusters give isolation.
Platforms give control.
The Missing Piece

Introducing MTO

The enterprise platform layer between Kubernetes and your teams

Your Teams
Team A
isolated tenant
Team B
isolated tenant
Team C
isolated tenant
Team N+
isolated tenant
↕
Platform Layer
Multi-Tenant Operator
Red Hat Certified
Tenancy Templates Hibernation FinOps Extensions Compliance
↕
AKS
OpenShift
EKS
Any K8s
Kubernetes Infrastructure
The platform you end up building anyway —
out of the box.

Everything Required for an
Enterprise Platform

🏢
Tenancy
True tenant abstraction — not just namespaces. Isolation across workloads, network, and storage. Enforced automatically.
📋
Templates
Golden templates for apps and infra. Standardization without slowing teams down. Controlled self-service.
💤
Hibernation
Auto sleep/wake for non-production workloads. Up to 60% cost savings. On-demand activation.
💰
FinOps
Per-tenant cost visibility. Showback and chargeback. Budgets, alerts, rate plans. No more cost blindness.
🔌
Extensions
Plug-and-play ecosystem: ArgoCD, Vault/OpenBao, Keycloak, observability stack. One platform, unified experience.
🛡️
Compliance
Policy-as-Code via Kyverno. Continuous enforcement — not point-in-time audits. Audit-ready by default. Governance without blocking developers.

Not features — capabilities required to operate Kubernetes at enterprise scale.

Pillar 1 of 6  ·  🏢

Tenancy

True tenant abstraction — not just namespaces

The Problem Today
✕
A namespace is not isolation
It's a label. Without enforced boundaries, teams can accidentally reach each other's workloads — and you won't know until something breaks.
✕
Manual setup for every team
RBAC, quotas, network policies — configured by hand, every time. One missed setting is a compliance gap.
✕
Auditors ask who has access to what
Misconfigured RBAC causes incidents. When the question comes, you don't have a clean answer — just a lot of manual digging.
With MTO
✓
Tenant as a first-class resource
Not just a namespace — a governed unit with its own identity, policies, and enforced boundaries from the moment it's created.
✓
Isolation provisioned automatically
RBAC, network policies, resource quotas — set up when a tenant is onboarded. No manual steps. No forgotten configs. No gaps.
✓
Real boundaries — workloads, network, storage
Enforced continuously, not just at setup. Teams can't see or affect each other. Security posture is consistent across every tenant.
Outcome
Secure multi-tenancy at scale — enforce once, inherit everywhere.
Pillar 2 of 6  ·  📋

Templates

Standardization without slowing teams down

The Problem Today
✕
Every team builds differently
No standardization means no repeatable baseline. Security policies, ingress configs, monitoring — each team reinvents them their own way.
✕
Onboarding a new team takes days
Platform team creates namespaces, applies RBAC, configures ingress, sets up monitoring — manually, every single time.
✕
Non-compliant configs reach production
Without enforced templates, teams ship whatever works locally. The platform team reviews and fixes — endlessly.
With MTO
✓
Golden templates for namespaces, apps, infra
Encode your platform's best practices once. Every team inherits them automatically — no negotiation, no exceptions.
✓
Self-service with built-in guardrails
Teams provision their own environments. Compliance is default behavior — not a gate they have to get through.
✓
Version-controlled, centrally managed
Templates evolve with your platform. Push an update once — all tenants inherit it. No drift, no one-off fixes.
Outcome
Teams onboard in minutes. Environments are consistent. Platform team stops firefighting.
Pillar 3 of 6  ·  💤

Hibernation

Stop paying for idle workloads

The Problem Today
✕
Dev and staging run 24/7
Nobody is using non-production environments at 2am, on weekends, or during holidays. But the compute bill doesn't know that.
✕
No automatic off switch
Cloud costs grow without discipline. Every idle pod is waste — but nobody owns the problem or has the tools to fix it.
✕
Finance can't explain the bill
"Cloud costs went up" is not a root cause. Without visibility into idle spend, every budget conversation is guesswork.
With MTO
✓
Automatic sleep / wake scheduling
Define business hours per tenant or namespace. Workloads spin down at night, back up in the morning — automatically.
✓
On-demand activation when teams need it
Teams can wake their environment instantly when they need it outside of schedule. No waiting, no platform team involved.
✓
Fine-grained control per environment
Different schedules for dev, staging, and test. Each team's workload profile is different — MTO respects that.
Outcome
Up to 60% savings on non-production. Immediate, measurable ROI from week one.
Pillar 4 of 6  ·  💰

FinOps

From cost blindness to cost control

The Problem Today
✕
One bill. Zero accountability.
The cluster invoice is a black box. You know the total. You don't know which team consumed what — or why it keeps growing.
✕
No per-team visibility
Engineering managers can't hold teams accountable. Nobody knows if their services are efficient. Budget overruns have no identifiable cause.
✕
Platform investment is hard to justify
Finance sees a growing cloud line item. Engineering can't defend it. The ROI of the platform is invisible — until it isn't.
With MTO
✓
Per-tenant cost visibility out of the box
Every tenant's consumption tracked automatically. No custom tooling. No manual aggregation. The data is just there.
✓
Showback and chargeback models
Report costs by team, project, or business unit. Integrate with FinOps tools and ERP systems. Finance finally has answers.
✓
Budget limits, alerts, and rate plans
Set spending thresholds per tenant. Get notified before overruns happen, not after. Cost accountability becomes the default.
Outcome
Cost is transparent, controllable, and accountable. Platform spend becomes defensible.
Pillar 5 of 6  ·  🔌

Extensions

One platform. Unified ecosystem.

The Problem Today
✕
Every team wires their toolchain differently
ArgoCD, Vault, Grafana, Keycloak — each team has their own setup, their own versions, their own break points.
✕
Inconsistent platform experience
A new engineer joins. Which team's way of doing things is right? There is no answer — just tribal knowledge and onboarding pain.
✕
Every integration is a support burden
Non-standard setups break in non-standard ways. The platform team inherits every problem — and there's no shared fix.
With MTO
✓
Plug-and-play enterprise integrations
ArgoCD, OpenBao/Vault, Keycloak, LGTM observability — standardized, tested, and provisioned per tenant on onboarding.
✓
Full stack delivered on tenant creation
A new tenant gets the complete platform — GitOps, secrets, identity, observability — not just raw Kubernetes primitives.
✓
Extensible without rearchitecting
Add new integrations as your platform evolves. The model is open — you're not locked into what ships out of the box.
Outcome
Consistent platform experience. Teams get the full stack from day one — no wiring required.
Pillar 6 of 6  ·  🛡️

Compliance

Built-in governance. Audit-ready by default.

The Problem Today
✕
Policies exist in documents, not in enforcement
Security baselines are written down. What actually runs in the cluster is a different story. The gap accumulates silently.
✕
Audit prep is a project in itself
Every audit cycle means weeks of evidence collection, gap analysis, and emergency patching. It's reactive, expensive, and stressful.
✕
Policy drift is invisible until it isn't
Configurations change over time. Policies become outdated. Nobody notices — until a security incident or a failed audit exposes the gap.
With MTO
✓
Policy-as-Code via Kyverno — always on
Policies aren't suggestions — they're enforced at admission. Non-compliant configs are rejected before they reach the cluster.
✓
Every tenant starts compliant
Compliance boundaries are set at tenant creation — ISO, DORA, SOC 2 baselines applied by default. No remediation, no exceptions to track.
✓
Built-in audit trail — always current
Who provisioned what, when, with which policies. Evidence is generated automatically — not collected manually three weeks before an audit.
Outcome
Audit-ready by default. Compliance is automatic — not a project that runs every quarter.

From Chaos
to Platform

Without MTO
✕
Cluster per team
Sprawl inevitable — complexity multiplies with every team
✕
Manual onboarding
Days per team — platform team is the bottleneck
✕
Security drift
Every cluster different — policies diverge over time
✕
Cost blindness
No visibility per team — finance can't explain the bill
✕
Platform team burns out
Queue grows — scaling headcount is the only answer
✕
Manual compliance
Audit prep is a project — painful every cycle
With MTO
✓
Tenants in shared clusters
Controlled scale — add teams without adding clusters
✓
Automated provisioning
Minutes per tenant — teams self-serve inside guardrails
✓
Policy-driven enforcement
Security inherits everywhere — no drift, no exceptions
✓
Per-tenant cost visibility
Showback from day one — every team accountable
✓
Self-service platform
Platform team scales without headcount — queue disappears
✓
Continuous enforcement
Audit-ready by default — governance is automatic

Why Not
Build It Ourselves?

Every team that goes down this road builds the same platform. Here's what that actually requires:

Multi-tenancy layer
3–4 months engineering
✓ Day one
Policy engine
2–3 months, ongoing tuning
✓ Day one
Cost tracking
3–6 months integration
✓ Day one
Templates system
2–3 months + drift maintenance
✓ Day one
Ecosystem integrations
Ongoing — every tool is custom
✓ Plug-and-play
Compliance framework
6–12 months, auditor-dependent
✓ Policy-as-Code, built in
Total DIY: 6–18 months of engineering. Then you own it forever. Key-person dependency. No roadmap. No support.

The hidden truth

You're not choosing between "buy MTO" or "build nothing." You're choosing between MTO and 18 months of internal engineering that produces an incomplete, unmaintained version of MTO.

MTO is not a cost.
It replaces internal engineering spend, accelerates delivery, and removes ongoing maintenance ownership.
"You're not avoiding the work.
You're choosing to own it forever."

What This
Enables

⚡
Faster team onboarding
From days to minutes. Teams provision their own environments inside compliance guardrails. No queue.
📉
Reduced operational overhead
Platform team stops firefighting. Policy enforcement is automatic. Hibernate non-prod for immediate cost reduction.
🛡️
Built-in compliance
ISO, DORA, SOC 2 — governance baked in from day one. Audits become a formality, not a project.
📊
Predictable cost model
Per-tenant visibility means accountability. Finance can justify platform investment. Budget overruns have a cause.

Platform becomes an enabler, not a bottleneck

FOR ENGINEERING LEADERSHIP
Deliver more with the same platform team. Stop reinventing governance. Spend engineering cycles on product.
FOR FINANCE & PROCUREMENT
Replace hidden engineering spend with a known, supported product cost. ROI is immediate via Hibernation savings alone.
FOR SECURITY & COMPLIANCE
Policy enforcement is continuous, not periodic. Audit evidence is generated automatically. No manual evidence collection.

How You
Start

No big-bang migration. No rearchitecting. Start where you are.

1
Pilot
Single cluster.
2–3 teams as tenants.
MTO deployed alongside
existing workloads.

Timeline: 1–2 weeks
2
Expand
Roll out templates
across all teams.
Enable Hibernation
on non-prod.
Activate FinOps.

Timeline: 4–6 weeks
3
Standardize
Compliance pillar live.
Extensions integrated
(ArgoCD, Vault, etc.).
Chargeback model
operational.

Timeline: 8–12 weeks
4
Scale
Organization-wide
platform.
Multi-cluster if needed.
Platform team as
internal product team.

Platform at scale
You don't need to transform everything.
You just need to start.

Why Stakater

🏅
Red Hat Premier Partner
Deepest OpenShift integration & support
✅
Red Hat Certified Operator
MTO is certified — enterprise-grade, supported lifecycle
🧠
Kubernetes Multi-Tenancy Specialists
This is what we do — not a side product
🔧
Product + Consulting
We deploy with you, not just ship a license

What customers say

"We reduced team onboarding from weeks to hours."

"Our platform team stopped being a bottleneck overnight."

"We finally have cost visibility by team. Finance is satisfied."

"We stopped spinning up unnecessary clusters."

The Decision

You will build this
platform layer eventually.

OPTION A
Build it yourself
6–18 months engineering
Ongoing maintenance forever
Key-person dependency
No roadmap, no support
Still incomplete at v1
OPTION B
Start with MTO today
Production-ready in weeks
Proven patterns, continuous updates
Supported by experts
Red Hat certified
Immediate cost savings (Hibernation)

Let's talk about a pilot.

Workshop / Assessment Pilot Deployment Platform Maturity Roadmap