24/7 SRE cover for AWS, Azure and GCP
We answer every alert, so your team only hears about the new ones.
Our in-house engineers cover your cloud production across all 21 shifts a week. One acknowledges every alert within 15 minutes and works out the cause and severity within 30, both written into your contract. Known issues are fixed by runbooks you approve. New issues reach your on-call engineer with the evidence already gathered.
- To acknowledge
- 15 min
- Every alert, in your contract
- Cause and severity
- 30 min
- Diagnosis, not a fix-time promise
- Shifts a week
- 21
- Nights, weekends and holidays
- Closed without paging you
- [X]%
- [Average across customers, last N months]
Answering alerts is half the job.
Cover · every hour
An engineer on every alert.
We start from the alerts your tools already send, so there’s nothing to install. Our engineers learn your systems and past incidents, then answer where your team already works.
- A likely cause for every alert, with the evidence
- A written report on what contributed to every incident, not who to blame
- Answers in your Slack channel and your ticket
Works with
Reliability work · every quarter
One or two goals a quarter, each with a target and a date.
We agree them with you, and your monthly report shows progress.
Cut the pages nobody needed to act on to 23% or lower by 31 December.
- July
- 38%
- Now
- 29%
- Target
- 23%
On track
Who answers at 3am
Named engineers, and a rota you can check.
You should know who is looking after your production before you sign. Here is who leads the team, how the rota works and how many customers each engineer covers at once.
Ken Tan
Founder, [title]
[One line: previous SRE roles, years on call, the kind of systems run]
- Engineers on the rota
- [N]
- All in-house, all write code
- Customers per engineer on shift
- Up to [N]
- A cap we hold, shown in your monthly report
- Where shifts are worked
- [Singapore and other locations]
- Daytime shifts where possible, so nobody works every night
- On every shift
- A primary and a backup
- [Confirm this is true before publishing]
Why the price works
A whole SRE rota for less than one SRE hire.
Covering nights in-house takes at least eight on-call engineers once you allow for holidays, sick days and people leaving. Our 24/7 on-call guide shows the arithmetic.
We can start at S$1,500 a month because AI investigates every alert before an engineer picks it up. That lets each engineer on shift cover up to [N] customers and still meet the 15-minute window. We cap that number, and your monthly report shows it.
Cost of 24/7 cover, per year
S$[amount]
S$18,000
Customers
What changed for the teams we cover.
Published with each customer’s permission, with figures taken from their monthly reports.
-
Placeholder: customer logo
[Quote from the customer’s engineering lead, cleared for publication]
[Name, role, company]
- Pages to their team each month
- [Before] to [after]
-
Placeholder: customer logo
[Quote from a second customer, cleared for publication]
[Name, role, company]
- Incidents users spotted first
- [Before] to [after]
Free alert audit
Send us a month of alerts. We will show you what we would have closed.
No access to your cloud needed. Export your alert history from PagerDuty, Datadog, Grafana or your cloud’s own monitoring. Within [5] working days you get a written report: how much of it is noise, which alerts fire together, which user journeys have no alert, and which issues a runbook could fix.
Your options
Three ways to stop waking your engineers, and where we fit.
| Topic | Hire SREs | AI SRE tool | Cloud MSP | Tracefox |
|---|---|---|---|---|
| Who answers at 3am | Your own engineer | Software investigates; your engineer still gets paged | A service desk, often shared across many customers | An in-house Tracefox SRE |
| Response time in a contract | Your own policy | Usually none for people | Varies by priority level | 15 min to acknowledge, 30 min to cause |
| Work on why alerts happen | If you staff for it | Suggestions only | Usually a separate project | One or two goals every quarter, in the fee |
| Cost | At least eight salaries | Licence by seat or usage | Custom quote, often a share of cloud spend | From S$1,500 a month |
See the full comparison, including when another option suits you better
One flat monthly fee, published.
Every plan includes 24/7 cover with both response times in your contract, quarterly reliability goals, a monthly report and a quarterly review. No licence markups and no agent fees.
Billed monthly; after the first three months, either side can end it with 30 days’ notice.
Flat monthly fee
- Up to 5 production services
- S$1,500a month
- 6 to 10 production services
- S$[price]a month
- More than 10 services
- Quotedon the 20-minute call