Contact
Talk to an engineer, not a salesperson.
Tell us what hurts. The first conversation is with someone who would actually do the work — no qualification call, no discovery deck. We reply within one business day.
Prefer to skip the form?
Two time zones, so someone is awake and at a desk — not on call from bed.
Send us a message
This opens your own mail app with everything filled in — nothing is sent until you press send there. Fields marked with an asterisk are required.
If you are already a client
Our response commitments.
Published, so you never have to ask. If you are already a client and need to raise an incident, email getintouch@aaira.techwith the severity in the subject line. These are the response targets attached to each severity.
P1 — Critical
Production down or unusable; confirmed security incident
- Response
- 15 min
- Hours
- 24×7
P2 — High
Major function degraded, or a painful workaround in place
- Response
- 1 hr
- Hours
- 24×7
P3 — Medium
Minor degradation or a fault affecting one user
- Response
- 4 hrs
- Hours
- Business
P4 — Low
Question, change request or scheduled work
- Response
- 1 business day
- Hours
- Business
Targets are measured from the moment an alert or ticket is raised, not from the start of the next business day. Severity is assessed against the impact on your users, and we will tell you when we disagree with a classification rather than quietly downgrading it.
Before you write
The questions we get asked first.
If your question is answered here, you do not need to send an email to find out.
How long does it take you to take over an existing environment?
Discovery and dependency mapping run in the first one to two weeks, a documented estate follows in three to four, and full round-the-clock cover typically starts between weeks six and eight. Nothing is put under management before it has been documented — inheriting an estate nobody has written down is how providers end up guessing at three in the morning.
Do we have to migrate to a different cloud to work with you?
No. We run AWS, Azure, Microsoft 365 and hybrid estates as they are. A migration only happens when there is a reason for it that survives a business case, and the disposition for every workload is written down and agreed before anything moves.
What happens to the tooling we already pay for?
It stays unless it is genuinely redundant. We instrument through open standards rather than a proprietary agent, so your telemetry remains yours and portable. Where we do replace something, the reason and the saving are in the monthly review.
What happens to our own engineers?
They stop carrying the pager for incidents they have already solved. Nothing in the model does architecture, design or judgement calls — those stay in-house. In practice the engineers who would have taken the 3am pages are the strongest internal advocates for the arrangement.
What if we want to leave?
Your infrastructure lives in your repository in Terraform or CDK throughout, changed by pull request, and your telemetry is in your own tenant. There is no proprietary layer to unpick. When the engagement ends you keep a platform you can run — that is the point of doing it as code rather than by clicking around a console.
Will you work with our auditors?
Yes. Every agent action is logged with its reasoning and attributed to a named agent identity, so it is distinguishable from human activity in your existing cloud audit trail, and it is exportable to your own SIEM. Quarterly restore tests are run with the evidence sent to you rather than described in a document.
Is the 24x7 cover real, or an on-call rota?
It is staffed from Philadelphia and Hyderabad, which is why two offices exist in those particular places — the follow-the-sun split means neither team works nights to give you daytime cover.
What is the smallest way to start?
A two-week health check with read-only access and no commitment. You get a written findings report covering security posture, resilience gaps and a costed list of savings visible in your own billing data. If you stop there, you keep the report.