Skip to content
Check outages

For Australian electricity distributors

Answer every customer at once — even when everyone calls at once.

We take your outage data to your customers across every channel — web, chat assistant and phone — answered by AI voice agents that scale with demand, so no one waits in a queue, and every byte can stay inside Australia.

Scale

Tens of thousands of simultaneous conversations

We manage and deploy the capacity, so you never have to. It scales on demand with the event — no queue, no hold music, and no abandoned calls on your worst night of the year.

Resilience

No single point of failure

Redundancy is built into our foundations, never bolted on afterwards. It is designed in at every layer — carrier, cloud, region and model provider — so one provider's bad night never becomes yours.

Sovereignty

Your data stays in Australia

Deployed data-sovereign first. Every part of the process — outage data, call handling, transcripts and recordings — is processed and stored in Australian regions only, under Australian jurisdiction. Nothing leaves the country. Zero-retention plans available.

Or hear it now — +61 3 7068 5441 answered by the voice agent, no queue.

Where things stand, plainly: the voice agent, the national outage feed and the public service are built and running today. Multi-cloud and multi-carrier failover and the fully onshore deployment are provisioned per distributor as part of onboarding, sized to your contracted capacity — we do not pre-run another network's redundancy, and we will not claim yours is live before it is.

The storm problem

On an ordinary day your outage line is quiet. During a severe weather event it can take a year's worth of calls in an afternoon — and nearly every caller is asking a version of the same three questions: Is it out? Why? When will it be back?

Contact centres can't be sized for that. Staffing for the worst night of the year means paying for idle capacity the other 360 days, so the sensible commercial decision is to under-provision — and accept that people wait on hold when they're frightened and in the dark.

That isn't an operational failure. It's a structural mismatch between fixed capacity and spiky demand — and it's exactly the kind of problem software solves better than rostering.

What that costs you

  • Customer frustration concentrated at your worst possible moment
  • Complaint and ombudsman exposure following major events
  • Staff handling repetitive status calls instead of genuine emergencies
  • Peak capacity paid for year-round and used a handful of days

This is already on the public record

In February 2024 a storm took 255,000 customers of a single Victorian distributor off supply. Its outage tracker crashed in the afternoon and stayed down for eight days. Customers did the obvious thing and picked up the phone.

255,000

customers off supply

43,000

calls on day one

79%

of calls not resolved by the automated menu were abandoned

22 min

average wait to reach a person

+198

further agents the distributor estimated it needed that day

$12M

court-enforceable undertaking that followed

The independent review the distributor commissioned found there was no feasible way to serve that level of demand. It also recorded that the distributor had no live chat or automated assistant available as an alternative channel, and that long waits meant not every customer on the life-support register was able to get through.

The regulator's finding is the part worth sitting with. In the Essential Services Commission's published words — "[the distributor] has acknowledged its failure to provide timely service interruption information to customers was a contravention of Victoria's energy laws."

The enforcement action was not about the storm, or how long restoration took. It was about not being able to tell people what was happening.

The channel nobody stress-tests

The commission reviewed all five Victorian distributors in December 2024 and found rigorous annual load-testing of outage tracker websites against that February peak — one to four and six times peak load, another to 5.37×, two more to 2.64×, another to 250,000 published outages.

4–6×

website, tested annually

1×

phone line — the channel the rules actually mandate

A 24-hour telephone service with a live operator is required under the Electricity Distribution Code of Practice. It is also the one channel whose capacity is set by a roster rather than by architecture — which is why nobody publishes a multiple for it. There isn't one to publish.

We have not named the distributor. The lesson belongs to the sector, not to one operator, and it could have been almost any network on the wrong night. Full sources are at the foot of this page.

One subscription. Three channels.

You send us outage information once. We make sure your customers can reach it whichever way they choose to ask — and that they get the same answer every time.

WEB

Your data, presented clearly

Your outages on a national map, attributed to you and linking back to your own page — so customers who don't know who supplies them still end up reading your information.

CHAT

An assistant that answers

Customers ask in plain language — "is my power out in Richmond?" — and get a direct answer drawn from your feed, at any hour, without a form.

PHONE

Your outage line, answered

Route your outage number to us. AI voice agents answer every call about your network using your data — during storms only, or permanently.

Built for the night everything goes wrong

A service that answers emergency-adjacent calls has to survive the same storm your network is fighting — and it has to be somewhere you're comfortable putting your customers' data.

Redundancy across providers

Provisioned at onboarding
  • Multi-cloud. If one major cloud provider has an outage, we fail over to another. Your outage line shouldn't depend on somebody else's region staying up.
  • Multi-carrier. If one telephony provider goes down, calls route via another.
  • No shared fate with you. We run independently of your contact centre, so an incident at your end doesn't take the public-facing answer down with it.

Australian data, kept in Australia

Deployment option

For distributors with sovereignty obligations — or who simply prefer it — we can deploy the whole pipeline inside Australian regions, so customer information is never processed or stored offshore:

  • Outage data ingested, stored and served from Australian regions only
  • Calls answered and transcribed onshore — recordings never leave the country
  • Address matching against Australian data, held here
  • No offshore support access to customer data
  • Zero-retention option: transcripts and recordings discarded after the call

Our own aggregation already runs in Sydney. A fully onshore deployment is configured per distributor and written into the agreement, so it's a commitment you can point at rather than a claim on a web page.

Three ways in

Capacity is the thing you have to size, so we lead with it. An agent is one AI voice agent holding one conversation — the same unit you already roster your contact centre in. Your included agents answer around the clock; burst agents are the surge capacity we contract to and load-test against, for the afternoon everyone calls at once. Exceed your burst and we bill the overage — we will not hand your customers a busy tone to protect our margin.

Start here

Any Day

Single agent · 10 burst agents

99.5% availability target

  • Activates on a volume threshold or your manual trigger
  • Two carriers, single cloud region
  • Annual load test to your contracted burst

Indicative pricing on request

Most distributors settle here

Rainy Day

5 agents · 50 burst agents

99.95% availability target

  • Takes every inbound outage-status call, warm-transfers the rest to you
  • Multi-cloud active-active, two carriers
  • Quarterly load test — results written up so you can hand them to a regulator
  • Service credits attached to the availability figure

Indicative pricing on request

For networks under a live undertaking

Stormy Day

20 agents · 100 reserved · 200 burst agents

99.99% availability target

  • Reserved capacity, not best-effort
  • Fully onshore data path, dedicated instance, isolated quotas
  • Monthly load tests plus an annual joint exercise with your emergency management team
  • Post-event reporting pack built for regulatory submission

Indicative pricing on request

The per-minute rate falls as the tier rises. You are paying for guaranteed capacity and availability, which is what actually costs us money — the minutes themselves are cheap to serve, so we price them that way.

How it works

  1. 01

    You send a feed

    Whatever format suits you — an API, a file drop, an existing endpoint. We normalise it.

  2. 02

    We publish it everywhere

    Web, chat and voice all answer from the same data, attributed to you and linking back.

  3. 03

    You route your calls

    Point your outage line at us during events or permanently. Entirely your decision.

You stay the source of truth

We never present ourselves as the authority on your network. Every outage links back to you, we show when data was last updated, and we don't display stale information as live.

Emergencies always escalate

Life support customers, fallen powerlines and anything safety-related are directed straight to you and to 000. The AI handles status questions, not emergencies.

Free for your customers, always. Members of the public are never charged to check an outage, use the chat assistant or call the number. The subscription sits with the distributor, not the customer.

Start a conversation

We're early, and we'd rather build this with distributors than around them. Tell us what your outage line looks like on a bad night and we'll show you what we can take off it.

[email protected]

We'll always correct or remove anything relating to your network on request.

Sources and attribution

Figures for the February 2024 event come from an independent post-incident review commissioned and published by the distributor concerned (July 2024), and from the Essential Services Commission's published findings on the resulting court-enforceable undertaking. Load-testing figures come from the commission's December 2024 review of Victorian distributors' emergency preparedness. All of it is on the public record and we will supply the full references on request. We have chosen not to name the distributor.

Essential Services Commission material is reproduced under CC BY 4.0; the quoted sentence above is altered only to remove the distributor's name, marked with square brackets. Findings from the independent review are summarised in our own words rather than reproduced.

We are not affiliated with, endorsed by, or acting for any electricity distributor, the Essential Services Commission, the Australian Energy Regulator, or the authors of any report cited here. Nothing on this page should be read as suggesting otherwise, and nothing here is a claim that our service would have altered the outcome of any past event.

Capacity figures and availability targets describe what we contract to, not measured historical performance. Pricing is indicative and subject to scope.