Frequently asked questions
Fifteen of the questions we hear most often, answered in two to four sentences each. If you want a single page to skim before signing up, this is it.
Summary
AlwaysOnIT is an AI-only IT support service for small and medium UK and European businesses. Every support ticket is handled by an AI agent over chat — there are no human technicians, no phone queues, and no on-site visits. It is designed for companies of roughly 5 to 50 staff that run on Microsoft 365 or Google Workspace and want immediate, predictable IT support without a traditional managed service provider (MSP).
AI IT support is a helpdesk service where an autonomous AI agent diagnoses and resolves common IT issues directly in chat, instead of routing tickets to a human technician. The agent uses a fixed set of tested procedures and scoped tools (Microsoft 365, Google Workspace, Intune, Jamf, Okta, Entra) so the same issue is fixed the same way every time, 24 hours a day.
A traditional MSP staffs a rota of human engineers and bills for their time. AlwaysOnIT replaces that helpdesk layer with a single AI agent that responds in seconds, 24/7, at one flat per-user price. There are no callbacks, no out-of-hours surcharges, no minimum-hours contracts, and no on-site visits. The trade-off is honest scope: AlwaysOnIT only handles what it can complete safely from chat, and clearly declines anything that needs physical presence or regulated accreditation.
AlwaysOnIT costs £4 per user per month (USD $5 / EUR €5) with a £12 minimum charge per company. Billing is monthly in advance with no contracts and cancellation at any time. Companies with fewer than three users pay the £12 minimum; above that, billing is simply £4 per active user. VAT is added where required.
There is no free trial and no setup fee. Pricing starts at the £12 minimum per company per month, billed monthly with no contract. Because there is no commitment, customers typically subscribe, raise their first ticket, and cancel if it is not a fit — there is no long onboarding to sit through first.
AlwaysOnIT supports Microsoft 365 and Google Workspace (mail, calendar, sharing, licensing, admin), identity platforms (Entra, Okta), MFA and conditional access, password resets, Windows and macOS endpoint diagnostics, MDM (Intune, Jamf, Kandji, Mosyle), common SaaS account provisioning, phishing triage, and security hygiene. The full scope is published on the What we support page.
AlwaysOnIT does not support anything that requires physical presence (cabling, hardware repair, on-site networking, Wi-Fi coverage surveys), on-premise servers, regulated workloads (FCA-supervised, NHS clinical, defence), bespoke software development, contractual penalty-backed SLAs, or live ransomware incident response. These limits are listed in full on the What we don't support page.
Yes. AlwaysOnIT operates with a Cyber Essentials aligned posture, UK data residency on AWS London (eu-west-2), TLS 1.3 in transit, AES-256 at rest with per-tenant keys, and zero standing access — credentials are brokered per task using short-lived scoped tokens and then revoked. Every prompt, tool call, and credential use is written to an immutable audit log that customers can export at any time.
No. Customer prompts, tickets, and account data are never used to train any model. Inference runs against the customer's tenant only, with strict context boundaries, and the underlying model providers (Anthropic and OpenAI) are contracted with zero data-retention agreements.
For most small businesses under 50 staff, yes — AlwaysOnIT replaces the day-to-day helpdesk function (resets, MFA, mailbox issues, device diagnostics, onboarding and offboarding). It does not replace a CISO, a security operations centre, an on-site engineer, or regulated-industry compliance leadership. Companies with those needs typically use AlwaysOnIT for tier-1 and tier-2 support and keep a specialist for governance and on-site work.
Every message is answered in under 30 seconds, every time, including overnight and at weekends. There is no queue, no rota, and no callback — the agent picks up the moment a ticket is raised and resolves or escalates within the same conversation.
Every user verifies their email before sign-in. Users authenticate through the company's existing identity provider where applicable (Microsoft Entra, Google Workspace, Okta), and device posture is checked on every session. For higher-risk actions — password resets, MFA resets, off-boarding — additional confirmation is required through the registered admin contact before the agent will proceed.
No. Low-risk, reversible actions (token rotation, MFA re-enrolment on a verified account, a single mailbox permission fix) run directly. Anything destructive or wide-blast-radius — tenant-wide deletions, mass policy changes, backup destruction, encryption key changes, off-boarding — requires an in-band approval from a named admin on the customer side before it executes.
No. The agent is required to state plainly that it is an AI if asked, and it signs every message. There is no roleplaying as a named technician, no fake support persona, and no implication that a person is on shift. This is a deliberate product rule.
When a request is out of scope, the agent says so on the first message, names the category (scope, safety, or certainty), and gives a one-line reason. Where a related in-scope action exists it is offered as an alternative. There are no vague holding replies, no quiet drop-offs, and no transfer to a human technician — because there are no human technicians at AlwaysOnIT.
AlwaysOnIT serves businesses in the United Kingdom and across Europe. All customer data, logs, and backups are stored in UK regions on AWS London (eu-west-2). The service is delivered entirely in English.
Get started
No onboarding calls. No sales process. Connect your devices and start raising tickets in under ten minutes.