RMT Engineering Logo
Use case · Voice, web chat · Inbound

Cover the hours you don't staff

AI handles out-of-hours and spillover volume, resolves what it can, books callbacks with a real position and ETA, and logs everything for the morning.

Bring your operating hours and one escalation path. That is the whole configuration.

We reply within one business day. No newsletter.

Illustrative interface. Hours, counts and timings are sample data from a demo tenant, not a performance claim. The screen shows an OptiML coverage view: a week grid of who answers by hour, with AI covering nights and weekends and joining the team at busy periods, a callback booked at 22:41 showing its position in the virtual queue and a morning ETA, and overnight totals for conversations resolved, callbacks queued and calls escalated to the on-call line.

Voicemail is not a night shift

  1. Step one

    Set hours and thresholds

    Business hours, public holidays and per-queue overflow thresholds are all timezone-aware, evaluated against the caller's own timezone rather than your head office's.

  2. Step two

    Route to the agent, not the machine

    Out-of-hours and overflow contacts route to the AI agent instead of to voicemail. The same agent, the same knowledge, the same guardrails as the daytime line.

  3. Step three

    Resolve, or book a real slot

    It resolves what it can outright. What it can't, it captures: the request, the details and the consent. Then it books a callback into the virtual queue with a position and an ETA the customer can hold you to.

  4. Step four

    Hand the morning over

    Your team arrives to tickets with the transcript, the summary and the captured data attached, either in the native ticket workspace or synced into the ITSM you already run.

Your hours, and one escape hatch

The escalation path matters more than the hours. Everything else can be tuned in week two; the route for a genuine emergency has to be right on night one.

  • Your operating hours. Per queue, with the public-holiday calendar that applies to each one.
  • An escalation path for genuine emergencies. A number, a rota or an on-call queue — and the signals that should trigger it.

Decide what “urgent” means before you go live. At 03:00, with no supervisor to ask, the agent will follow the rule you give it exactly. Write the rule so that an ambiguous case escalates rather than waits.

Evenings, weekends, and the on-call line

A clinic covers evenings and weekends with an agent that answers service questions, books appointments against real availability, and routes anything clinically urgent to the on-call line immediately. Nobody hears a voicemail greeting, and nobody waits until Monday.

The five conditions the router evaluates, and what each one does with the contact.

Routing conditions, the destination, and what the customer gets
Condition Where it goes What the customer gets
In hours, queue under threshold The human queue, as today Nothing changes
In hours, queue over the overflow threshold The AI agent, per queue An answer now, instead of a hold tone
Outside business hours The AI agent Resolution, or a callback with a position and an ETA
Public holiday on that queue's calendar The AI agent The same as out-of-hours, without a rota change
An urgency signal you defined The escalation path, immediately A person, now, with the context captured so far

What the night leaves behind

Timezone and holiday aware

Call forwarding and calling rules follow the contact's own timezone and local public holidays, not your office calendar.

A position, not a promise

Callbacks go into a virtual queue with a real position and ETA, and land in the same inbox the team works from in the morning.

Tickets, not a voicemail list

Native tickets, or two-way sync with one of 11 ITSM connectors and a universal adapter for the tool nobody else connects to.

Reviewable in the morning

Every overnight conversation is scored on 15 dimensions, so an unattended shift is still an audited one.

The parts this runs on

This is the cheapest use case to start with, because it adds coverage rather than changing anything your team does between nine and six.

We publish outcome figures only with the customer's written approval, so this slot stays empty until one is signed off. In the meantime the mechanism above is the honest version.

Until thenBring your own scenario and we will run it live

Talk to our team

The platform underneath

Everything on this page is the OptiML CX Platform configured for After-hours & overflow, not a separate product. These are the modules it leans on.

Book a working session

Play us last night's voicemails

Send a handful of the messages your team picked up this morning. We ground an agent in your content and run those same calls through it. You see which ones would have closed overnight and which would have escalated.

  • 30 minutes
  • Your own overnight calls
  • No slide deck unless you want one

Book a demo

30 minutes, on your own content. No slide deck unless you want one.

or reach us directly

Your details stay private. We never share them.

This website uses cookies.

Cookies are small text files that allow us to create the best browsing experience for you on our site. By continuing to use this website or clicking "Accept & Close", you are agreeing to our use of cookies. To understand how we use cookies or how to manage them, please see our cookies policy.

Ask OptiML

Powered by RMT Engineering