Managed Services: AI on every alert, engineers behind every fix

AI takes the first pass at any hour. Our engineers confirm every finding and approve every change, and from Professional they apply the fix and harden it.

Request a quote Compare service levels

Choose a service level

What an hour of downtime costs you

EssentialsAn inconvenience
ProfessionalLost revenue
EnterpriseA loss at any hour

A small platform that earns most of your revenue still takes a higher level.

Starter

Watches and reports

$450+

a month for one Small platform, with no level fee

Request a Starter quote

Who it suits: One Small platform: a website, a small web app or an internal tool.
Reviews included: Not included.
First reply: 4 hours to a critical issue

Essentials
Watches and guides
AI removes the noise, and our engineers guide your team through each fix.
$900+
a month, with one Small platform
What changes at this level
Reviews included
Not included.
First reply
1 hour to a critical issue
Request an Essentials quote (opens in a new tab)
Who does what
What the AI does
Groups related alerts, drops duplicates and explains the likely cause.
What our engineers do
Confirm what the AI finds and guide your team through the fix.
What you get
Fewer, better alerts and a monthly report with recommendations.
You receive
- Alerts that need a person, sent to your contacts
- A monthly service and cost report
- Patch and backup status in that report
- A record of every change we make
How it runs
1. Each alertThe AI groups it and explains the likely cause, and we tell your contacts what it means.
2. Each monthThe service and cost report, with what we recommend.
3. On requestSmall changes, each recorded with who approved it.
Your part
- Apply the fixes we guide you through
- Approve maintenance windows and changes
- Own the application and its releases
- Keep the accounts, the root user and the data
Professional
Fixes and prevents repeats
Includes Essentials
AI drafts the fix, and our engineers apply it, test your restores and stop repeat incidents.
$1,800+
a month, with one Small platform
What changes at this level
Reviews included
A Health Review twice a year. It is $3,000+ on its own.Free
First reply
30 minutes to a critical issue
Request a Professional quote (opens in a new tab)
Who does what
What the AI does
Also drafts the fix, and runs runbooks you approved in writing on known, safe faults.
What our engineers do
Apply fixes, test restores and review costs each month. One named engineer.
What you get
Faster fixes, fewer repeat incidents and recovery you have tested.
You receive
- Everything in Essentials
- Dashboards for application and database health
- Restore test results
- A monthly cost review, with actions
- Notes from each architecture review
- A Health Review twice a year, with its report and 90-day plan
How it runs
1. Each alertThe AI drafts the investigation and the fix. We approve and apply it, and keep your contacts updated.
2. Each monthThe service report and the cost review, with actions agreed.
3. On a scheduleRestore tests and an architecture review.
4. Twice a yearA Health Review of your platforms, with its report and a 90-day plan.
Your part
- Approve changes and maintenance windows
- Decide which cost actions to take
- Own the application code, with our DevOps help
- Keep the accounts, the root user and the data
Enterprise
Owns the outcome
Includes Professional
We agree reliability targets, and an engineer owns each major incident, the drills and the change calendar.
$4,500+
a month, with one Small platform
What changes at this level
Reviews included
A Deep Dive twice a year. It is $6,500+ on its own.Free
First reply
15 minutes to a critical issue
Request an Enterprise quote (opens in a new tab)
Who does what
What the AI does
Also tracks the reliability targets we agree and flags a risk before one is missed.
What our engineers do
An incident lead, recovery drills, the change calendar and a quarterly review. A named engineer and a deputy.
What you get
Reliability targets you can count on, and the evidence for your audit calendar.
You receive
- Everything in Professional
- Service-level indicators, reported each month
- A written summary after each major incident
- Restore drill reports
- A written plan from each quarterly review
- A Deep Dive twice a year, with its report and roadmap
How it runs
1. Each alertOn-call response, with an incident lead for a major incident.
2. Each monthThe service report, the cost review and the service-level indicators.
3. Each quarterAn architecture review with your technical leads.
4. On your scheduleRestore drills, patch windows set in your change calendar, and a Deep Dive twice a year.
Your part
- Agree the service-level targets with us
- Join the quarterly review with your technical leads
- Approve changes through the change calendar
- Keep the accounts, the root user and the data

Starting prices for one Small platform. First replies run from the alert, at any hour. A first reply is not a fix.

How we get access

Free reviews with Professional and Enterprise

Professional includes a Health Review twice a year ($3,000+ on its own), and Enterprise a Deep Dive ($6,500+). No charge, for the platforms we run for you: up to three workloads for the Health Review and three accounts for the Deep Dive.

What each review covers

From alert to fix, at any hour

The AI is first on every alert. Engineers do the work that keeps it fixed.

A screen with dashboards of charts and quality scores

The AI, at any hour
A named engineer approves
Your named engineers

The AI, at any hour

Alert

An alarm fires in your account and reaches us and your own contacts.

The AI, at any hour

Group

Related alerts become one incident, and duplicates are dropped.

The AI, at any hour

Investigate

Logs and recent releases are read, and a likely cause is drafted.

The AI, at any hour

Propose

A fix is proposed, with what it changes and how to undo it.

Your named engineers

Confirm

An engineer checks the finding against how your system really runs, and sets the severity.

Your named engineers

Fix

An engineer designs the fix, and from Professional applies it through a role you can revoke.

Your named engineers

Harden

The weak spot is closed for good: a design fix, a new alarm, a restore test.

Your named engineers

Report

You get the record. Enterprise adds a written summary.

Where the work goes

Illustration

A resilient platform
A less resilient platform

The AIEngineers

The AI takes the routine work. Engineers take what it cannot, so the less resilient the system, the more of the work is theirs.

From Professional, our engineers apply the fix and harden it. At Starter and Essentials, they tell your team the fix to apply. The first pass runs on your cloud’s own agent: AWS DevOps Agent, Azure SRE Agent or Gemini Cloud Assist investigations, a preview.

How fast a person replies

SeverityWhat it meansStarterEssentialsProfessionalEnterprise

CriticalProduction is down.4 hours1 hour30 minutes15 minutes

HighA main function is badly degraded.12 hours4 hours2 hours1 hour

MediumDegraded, with a workaround.1 business day12 hours8 hours4 hours

LowA question or a small fault.3 business days2 business days1 business day1 business day

A first reply is not a fix.

Price by platform size and service level

Level feeThe engineers, the reports and the reviews

Each platformPriced by size: Small, Standard or Large

Your monthEach platform you add costs only its own price

Work out your price

Service levelStarterone Small platform, $450Essentials$650 level feeProfessional$1,300 level feeEnterprise$3,600 level fee

Your platforms

SmallUp to $3,000 a month, one region

$250 each

Standard$3,000 to $12,000 a month, up to two regions

$500 each

Large$12,000 to $40,000 a month, three or more regions

$1,000 each

More than 30 platforms of one size? We quote it against your estate.

Starting price

$900a month

Level fee: $650
Small: 1 × $250: $250
About per platform: $900

What counts as a platform: One production application, with its staging and test, in one cloud account. Its size is the highest limit it reaches, confirmed in the discovery week. Above $40,000 a month is quoted from the Large price.
Not in the price: The cloud bill itself. The final price is agreed in writing first.

What we run, area by area

Monitoring

AI watches health, capacity and cost, and cuts the noise so an alert means something.

What we do

Set a baseline of alarms for each resource type
Let the AI group related alerts into one incident
Tune or remove any alarm that needs no action

You receive

Each alarm, with an owner and a threshold
The alert history, in the monthly report

On each cloud

AWS: Amazon CloudWatch alarms and dashboards, with AWS Health events
Azure: Azure Monitor alerts and workbooks, with Azure Service Health
Google Cloud: Cloud Monitoring alerting and dashboards, in Google Cloud Observability

We scope each workload on your cloud.

Incident response

The AI starts the investigation, and an engineer confirms the cause and the fix and tells you what happened.

What we do

Start an AI investigation the moment an alert arrives
Confirm what is affected and set the severity
Apply the fix through a role you can revoke, from Professional

You receive

An incident record for each one
A written summary after a major incident, at Enterprise

On each cloud

AWS: AWS DevOps Agent, with CloudWatch alarms and Amazon EventBridge rules paging our on-call tool
Azure: Azure SRE Agent, with Azure Monitor alerts and action groups
Google Cloud: Gemini Cloud Assist investigations, a preview, with Cloud Monitoring alert policies

We scope each workload on your cloud.

Patching and security

We keep systems updated in windows you agree, and watch the findings your cloud raises.

What we do

Agree maintenance windows and patch baselines
Snapshot first, patch staging before production
Investigate failed updates and roll back when needed

You receive

A patch status report
Security findings, each with an owner and a fix

On each cloud

AWS: AWS Patch Manager, with AWS Security Hub, AWS Config and Amazon GuardDuty
Azure: Azure Update Manager, with Microsoft Defender for Cloud
Google Cloud: VM Manager OS patch management, with Security Command Center

We scope each workload on your cloud.

Backup and recovery

We keep copies you can restore, and prove that they work.

What we do

Set backup plans from the recovery targets you give us
Watch every backup job and fix the ones that fail
Restore from backup during an incident

You receive

Backup coverage and restore test results
A drill report against your targets, at Enterprise

On each cloud

AWS: AWS Backup, with its restore testing
Azure: Azure Backup and Azure Site Recovery
Google Cloud: Backup and DR Service

We scope each workload on your cloud.

Cost and capacity

We show where the money goes, take out waste and keep capacity ahead of demand.

What we do

Report spend by service and by tag each month
Flag idle and oversized resources
Switch non-production off outside working hours

You receive

A monthly service and cost report
A list of savings, with the effort for each, from Professional

On each cloud

AWS: AWS Cost Explorer, AWS Budgets and AWS Compute Optimizer
Azure: Azure Cost Management and Azure Advisor
Google Cloud: Cloud Billing reports and Active Assist recommendations

We scope each workload on your cloud.

Changes and releases

Every change goes through code and review, so it is recorded and can be reversed.

What we do

Take requests for scoped changes: an alarm, a backup plan
Make infrastructure changes through reviewed code
Support your release pipeline, from Professional

You receive

A record of each change
The templates, in your own repository

On each cloud

AWS: AWS CloudFormation or Terraform
Azure: Bicep or Terraform
Google Cloud: Terraform

We scope each workload on your cloud.

How a service starts

Step 1: DiscoverWe agree scope and contacts and read what exists.
Step 2: ConnectAlarms and backups are set, and the AI reads your telemetry.
Step 3: ShadowWe run beside your team. Nothing changes without approval.
Step 4: Go liveAlarms have owners and a restore has worked.
Step 5: RunReports and the reviews begin.

Customer work

All case studies

Questions before you request a quote

Do you need root access, and who owns the accounts?

No root user, ever. We work through roles you create and can revoke, and your cloud’s audit log records every action. Accounts, data and root access stay with you.

Does the AI change anything in our accounts?

Not on its own. It proposes and a named engineer approves, except a runbook you approved in writing for a known, safe fault, from Professional.

Why not just use the cloud’s own agents?

We use them for the first pass. They find and explain. Our engineers confirm each finding and answer for the result, and from Professional a named engineer fixes and hardens it.

Which clouds do you run workloads on?

Any cloud you run on. We scope each workload at the start, so you know what we cover.

How is it priced, and what if we change level or leave?

A level fee for the team and the reviews, plus a price for each platform by size. If you leave, you keep your accounts, data and templates, and we hand over the runbooks and remove our roles.

Managed Services: AI on every alert, engineers behind every fix

Choose a service level

Starter

Essentials

What changes at this level

Who does what

You receive

How it runs

Your part

Professional

What changes at this level

Who does what

You receive

How it runs

Your part

Enterprise

What changes at this level

Who does what

You receive

How it runs

Your part

Free reviews with Professional and Enterprise

From alert to fix, at any hour

Where the work goes

How fast a person replies

Price by platform size and service level

Work out your price

What we run, area by area

Monitoring

What we do

You receive

On each cloud

Incident response

What we do

You receive

On each cloud

Patching and security

What we do

You receive

On each cloud

Backup and recovery

What we do

You receive

On each cloud

Cost and capacity

What we do

You receive

On each cloud

Changes and releases

What we do

You receive

On each cloud

How a service starts

Customer work

Questions before you request a quote

The engineering behind it

What usually comes next

Tell us what you run and need next