What we support

The boring 95%. Handled.

If a competent helpdesk engineer could resolve it remotely, our AI almost certainly can. Here is the working scope.

Identity & email

  • ·Microsoft 365 / Entra ID
  • ·Google Workspace
  • ·Okta
  • ·MFA resets, conditional access, mailbox issues
  • ·Shared mailboxes, distribution lists, calendars

Endpoints

  • ·Windows 10 / 11 desktops & laptops
  • ·macOS (Apple Silicon & Intel)
  • ·Intune, Jamf, Kandji, Mosyle MDM
  • ·Patching, profile drift, login & sync issues
  • ·Onboarding & off-boarding flows

Productivity SaaS

  • ·Slack, Microsoft Teams, Zoom
  • ·Notion, Confluence, Asana, Linear, Jira
  • ·Adobe Creative Cloud licensing
  • ·1Password, Bitwarden
  • ·GitHub & GitLab seat management

Network & access

  • ·Wi-Fi & VPN client-side issues
  • ·Cloudflare Zero Trust, Tailscale
  • ·DNS records on managed registrars
  • ·Basic firewall rule reviews

Security hygiene

  • ·Phishing triage & user reports
  • ·Credential rotation after a leak
  • ·Cyber Essentials evidence gathering
  • ·Endpoint AV / EDR alert review

Procurement helpers

  • ·License right-sizing recommendations
  • ·SaaS renewal reminders
  • ·Hardware spec advice (you order)
  • ·Supplier ticket coordination
A note on edge cases — if a request sits at the boundary of our scope, the AI will say so plainly and tell you what it would need to proceed (a vendor escalation, an admin approval, or simply a polite "this isn't us").

See pricing, what we don't support, or the FAQ for related detail.

Issues we handle automatically

What the AI service does on its own.

The list below is what passes our decision checks every time: verified user, supported platform, online-only, low risk, high certainty. The AI service handles these directly in chat — diagnosing, guiding, and where it has the right tools, acting on your account.

Cloud & SaaS support

  • Microsoft 365 email issues

    Missing messages, send/receive failures, quarantine review, shared mailbox access, calendar permissions, signature and forwarding rules.

  • Teams and OneDrive problems

    Sign-in loops, sync conflicts, missing files, channel and SharePoint permission fixes, OneDrive storage and version recovery.

  • Password resets and account lockouts

    Verified resets on Microsoft 365, Google Workspace, Okta and Entra. Locked-out accounts unlocked once identity is confirmed.

  • Multi-factor authentication troubleshooting

    Re-enrolment after a lost or replaced device, authenticator app setup, sign-in failure diagnosis, conditional access review.

  • User onboarding and offboarding guidance

    Account creation, licence assignment, group membership, mailbox handover, data export, and clean account closure on supported platforms.

Endpoint diagnostics (remote only)

  • Windows and macOS performance checks

    Startup slowness, high memory or CPU use, storage pressure, background process review — all run remotely with your consent.

  • Update and patch diagnostics

    Why an update is failing, which patch is missing, whether a device is on a supported OS build, and how to bring it back into line.

  • Security hygiene checks

    Disk encryption status, screen lock policy, browser and OS update level, antivirus state, risky browser extensions.

  • Interpreting security alerts

    Plain-English explanation of EDR, M365 Defender or Google alerts: what triggered, how serious it is, and the safe next step.

Service behaviour

  • Step-by-step guidance

    Each fix is broken into numbered actions you can follow in your own time, with screenshots or commands where they help.

  • Clear explanation before any action

    Before the agent changes anything on your account or device, it states what it will do, why, and what the effect will be.

  • Safe, repeatable fixes

    The agent uses a fixed set of tested procedures. The same issue gets the same fix, in the same order, every time.

  • Confirmation of outcomes

    After each change, the agent checks the result, tells you what it found, and asks you to confirm before closing the ticket.

Frequently asked questions

Questions about what our AI IT support covers

Specifics on Microsoft 365, devices, accounts and security, beyond the category lists above.

What Microsoft 365 problems can the AI resolve?

The AI handles the full range of everyday Microsoft 365 support: Outlook profile and sync failures, shared and delegated mailboxes, distribution lists, email signatures, mail flow and delivery problems, Teams membership, meetings and call quality, SharePoint and OneDrive permissions and sync errors, licence assignment and reassignment, and admin centre configuration questions. It also covers the identity layer around 365 — password resets, multi-factor authentication enrolment and re-enrolment, and Conditional Access prompts that block a sign-in. These categories make up the large majority of SME IT tickets, which is exactly why an always-available AI agent is a good fit for them. Anything requiring physical access to hardware sits outside this scope.

Can the AI reset passwords and fix MFA lockouts?

Yes, and these are among the most common requests we handle. Password resets are offered as guided self-service where your tenant allows it, and as an admin-assisted reset once identity verification has completed where it does not. Multi-factor authentication is covered end to end: enrolling a new authenticator app, re-enrolling after a lost or replaced phone, clearing a stuck registration and explaining a Conditional Access prompt that is blocking sign-in. Because verification always runs before the change, a lockout can be resolved at two in the morning without weakening security. This is typically a five-minute conversation rather than a next-morning callback.

Does AlwaysOnIT support Windows and macOS devices?

Yes, for diagnostics and guided remediation on both platforms. That covers slow or failed startup, memory and disk pressure, storage cleanup, update failures, driver and peripheral issues, browser problems, application crashes and general performance troubleshooting. The AI walks you through the tools already built into Windows and macOS and interprets the output you report back, so no agent or remote-control software is installed on your machine. What it cannot do is anything physical: replacing a failed drive, repairing a screen, swapping RAM or dealing with a device that will not power on at all. In those cases it will write up a clear summary for a hardware engineer.

Can the AI help with phishing emails and security incidents?

Yes, within a clearly defined scope. The AI triages suspected phishing and spoofing, tells you whether a message is likely malicious and what the indicators are, and guides immediate containment steps — locking down an affected account, forcing a sign-out of active sessions, resetting credentials and checking for mailbox rules an attacker may have created. It also advises on Cyber Essentials alignment and general security hygiene such as MFA coverage, legacy authentication and admin account count. It is not a security operations centre: there is no 24/7 threat monitoring, no SIEM, no forensic investigation and no incident response retainer. For a confirmed breach, engage a specialist firm.

What SaaS applications beyond Microsoft 365 are covered?

The AI supports the common SaaS tools SMEs run alongside Microsoft 365 — the account, access and configuration layer rather than deep product administration. That means user provisioning and de-provisioning, sign-in and single sign-on problems, permission and sharing issues, licence questions and integration troubleshooting for widely used business applications. Because the AI reasons from the vendor's current documented behaviour rather than a fixed script, it copes well with tools our support team has never been trained on specifically. What it will not do is bespoke configuration, custom development or administration of specialist line-of-business software that requires vendor-certified access.

Is there a limit on how many tickets we can raise?

No. The £4 per user per month price covers unlimited support conversations for every licensed user, with no fair-use cap, no ticket bundle and no per-incident charge. This is a genuine structural difference from human-delivered support, where every ticket consumes technician time and pricing has to reflect that. An AI agent's capacity does not degrade with volume, so a month where everything breaks costs exactly the same as a quiet one. It also removes the incentive for staff to avoid raising small issues, which is usually how small problems become large ones.

What are the benefits of publishing your support scope openly?

Most disputes with IT providers come from a mismatch between what the customer assumed was covered and what the contract actually said. Publishing both what we support and what we do not, in plain English and before you sign up, removes that ambiguity entirely. You can decide in five minutes whether the service fits your business rather than discovering the boundary during an incident. It also keeps us honest: because the scope is public, we cannot quietly narrow it. If something you rely on is not listed, that is a clear signal to keep a hardware or on-site provider alongside us rather than expecting one supplier to do everything.

Get started

Open a chat. The AI picks up before the second ring — every time.

No onboarding calls. No sales process. Connect your devices and start raising tickets in under ten minutes.