Everything in the service desk

This is the whole product, described in terms of what it does on a Tuesday afternoon rather than what it sounds like in a comparison grid.

There is no separate "enterprise tier" hiding the parts that make it work. Ticketing, the help centre, email intake, the knowledge base and reporting are on every plan including the free one; the paid plans raise limits and add live chat, the API, the service catalog and, from Growth, the custom report builder.

Handling requests

The part your team lives in all day.

Tickets with a real lifecycle

Twelve statuses and a state machine that refuses a transition that makes no sense, so the queue cannot drift into a state nobody can explain. A resolved ticket can be reopened within seven days.

Internal notes kept visibly separate from replies the requester sees
Merge duplicates, link related, parent and child tickets
Major-incident flag, and a timeline of every change
Status choices that only ever offer the moves that are allowed

One queue, four channels

Email, the public help centre, live chat and the API all create the same kind of ticket, with the channel recorded. There is no second inbox to check.

Auto-categorisation from keyword rules you control
Ticket templates and canned replies with placeholders
Filters that survive in the URL you share

Email that threads correctly

Connect your existing helpdesk@ mailbox over IMAP. Replies attach to the right ticket using In-Reply-To and References, then the ticket number in the subject — and auto-replies are detected so you never end up in a loop with an out-of-office.

Acknowledgement sent from the mailbox that matches the queue
Per-message log that says why anything was skipped

Commitments and process

SLAs with real business hours

Response and resolution targets per priority, working hours per weekday, your holidays, and clocks that pause while you are waiting on the requester.

Warning at 75% and a breach event, both of which can drive automation
Every ticket shows each deadline and whether it is running, paused, met or missed
SLA state visible in lists and in every report

Built around ITIL 4 practices

Five registers for the practices a service desk runs: incident management, service request management, problem management with known errors, change enablement with approvals, and security incidents. Each has its own lifecycle, its own fields and its own links — there when an auditor asks, and out of the way when nobody is asking.

Change approvals, risk and impact, a planned date, CAB group and backout plan
Problems with a known-error database and linked incidents
Service requests raised from the catalog with an approval chain

Service catalog and approvals

Publish the things people can ask for, with a form for each, and route them to a manager or a named approver. Approvers are notified, decide from their Approvals page, and every decision is logged.

Deflection — fewer tickets, not faster ones

The cheapest ticket is the one nobody needed to raise.

A knowledge base people actually use

Articles with revisions, ratings and a public or internal audience. Guided walkthroughs take someone through a fix step by step and hand off to a pre-filled ticket when the self-help does not work — so a failed article still produces a good ticket instead of a bad one.

Draft an article from a resolved ticket in one click
Nightly mining proposes drafts from tickets you have already solved
Questions the chat assistant could not answer become a to-write list

Live chat with an assistant that checks the database

The assistant verifies who it is talking to against your employee directory, answers from your knowledge base with a link to the source, looks up real ticket status, and hands over to a person when it should.

Answers about your data come from the database, never from a language model
Identity verified by employee code and mobile, with an email OTP for anything sensitive
Agent console with presence, transfer, and convert-to-ticket

A help centre at supportit.in/yourcompany

No DNS record to add, no certificate to buy. Send your staff the link and they can raise a request, track it, search the knowledge base and start a chat.

Knowing how you are doing

Reports that answer the question you were asked

Ticket volume, open versus resolved, SLA performance, agent distribution, category and priority ageing, requester volume, chat deflection, first response and resolution times, CSAT and reopened-ticket analysis.

Pin any report to your dashboard
Export to CSV, Excel, Word or PDF
Schedule one to arrive by email daily, weekly or monthly

A report builder that cannot be injected

Build your own from allow-listed datasets and fields. User input never reaches SQL except through that allow-list — which is a security property, not a limitation we are apologising for.

Automation without a scripting language

Rules are an event, some conditions and some actions. Notify, escalate, set a status, add a note or send an email; webhooks go out separately, signed, for the same events. Every firing is logged with what it did, and loops are stopped at three levels.

Fitting into what you already have

REST API and webhooks

A documented API with bearer tokens or OAuth2, cursor pagination, idempotent writes and per-plan rate limits, plus outbound webhooks signed with HMAC and retried with backoff. The OpenAPI 3.1 description is in the product.

Import and export

Bring employees, service catalog items and knowledge-base articles in from CSV or Excel, with column mapping, a row-by-row error preview and a rollback. Take everything out again at any time, in any account state, at no charge.

Works on a phone properly

Installable as an app, with an offline page, a bottom navigation bar, tables that become card lists rather than sideways scrolling, and forms that do not make iOS zoom every time you tap a field.

Questions about how it works

Is it ITIL compliant?

It is built around ITIL 4 practices rather than certified against them. Incident management is on every plan, service requests with a catalog and approvals from Starter, and problems with known errors, changes with approvals and security incidents from Growth. Release management and a risk register are not part of this edition.

Do we have to use the AI features?

No. Out of the box the assistant runs on rules and your knowledge base, and no AI model is called — we do not run one for customers. To add AI answers and drafting, connect your own AI model in AI Studio: Ollama on your own server, or any OpenAI-compatible endpoint you choose.

Can requesters see each other's tickets?

No, and that is enforced at the query rather than by hiding a link. A requester sees only their own tickets and requests; an attempt to reach another's is refused.