Unified CCaaS platform Voice, chat, QA, and workforce management — unified. Local Phone #
Global DID New Local numbers, 150+ countriesChoose Your Model
Enhance Your Team
Answering Service
telecommunications
AI agents and human specialists handle service outages, plan changes, and billing questions in 65+ languages — keeping subscribers loyal across every channel.
The short version: put AI on outage status, data balances and plan lookups; put a person on bill disputes, porting, and any callback that crosses a timezone. AI voice is $0.15 a minute, chat $0.15 a session, a shared agent $300 a month for 500 minutes, half-seat minimum. Five days is our normal window to staff your account. And one thing to be blunt about: you are the one industry that can audit our carrier story, because it is your own trade.
how it plays out
Service outage during peak hours
Plan migration with number portability
Unexpected charge on monthly invoice
Family account management request
the team behind it
real members of the global team Telecom-Trained Specialists
65+ Languages
Follow-the-Sun Coverage
the part nobody prices in
Every support vendor selling to a global brand says some version of "we can give you a local number in your markets." Most buyers have no way to test that sentence. You do. You know that provisioning in most jurisdictions requires a local legal entity, proof of a local address, and either a telecom licence or a carrier relationship in that country — because getting those is what your own regulatory team does for a living.
So ask the question you would ask a wholesale partner. Not "do you support Germany" — "who holds the number, and what happens to it if we leave." A vendor reselling a single upstream supplier answers the first and goes quiet on the second. We will not tell you we already hold numbers everywhere.
The second thing that makes telecom different is the shape of the volume. Most industries have peaks you can forecast — a season, a promotion, a launch. Yours has a fibre cut. Four hundred people ask the same question inside an hour, none of them need a person for it, and every one calls back in twenty minutes if the first answer carried no restoration window.
That combination — one correct answer, enormous volume, an hour nobody staffed for — is the clearest case for automation on this site. It is also why the human half matters more here.
who takes which call
Most vendors sell you either an AI receptionist or a live answering service and leave the allocation to you. Since we run both, here is the split we would actually recommend — including the calls we think should never touch automation.
| The call | Route to | Cost of getting it wrong |
|---|---|---|
| "Is there an outage in my area?" | AI | A caller who reports a fault you already know about |
| "How much data is left on my plan?" | AI | A support minute spent on a lookup |
| Add a line, change plan, check upgrade eligibility | AI | An upgrade that stalls |
| "Why is my bill $40 higher than last month?" | Human | A billing dispute that becomes a chargeback |
| Third call this week about the same line | Human | A port-out you have already lost |
| Number porting — in or out | Human | A customer stranded between two carriers |
| A callback into a different timezone | Human | A legal exposure, not just a bad time to call |
what it costs
Most vendors in this category quote by form. These are our published packages — not a quote, since language mix, hours and volume all move the number. The per-minute figures are simply the monthly price divided by the included minutes.
| Option | Published price | Per minute | What it's for |
|---|---|---|---|
| Live shared team | from $650 / month | includes 1,500 min → ≈ $0.43/min | Overnight and weekend volume, when a truck roll is not happening anyway |
| AI voice | $0.15 | per minute | Outage status, restoration windows, "is it just me or the whole street" |
| Web chat | $0.15 | per session | Plan changes, add-a-line, data top-ups and device upgrade eligibility |
| Live shared team | from $300 / month | includes 500 min → ≈ $0.60/min | Bill explanations — the proration month after any plan change |
| Dedicated agent | from $5.80 | per hour, per agent | One team that knows your plan matrix, your proration rules and which outages are already open |
Anything involving a person is a starting price. Two things move it: which languages you need, and how senior an agent the work actually requires — a bilingual agent handling refund authority is not the same hire as one taking bookings. AI voice and web chat are flat, because there is no hiring in them. And every model above is open to every business on this page: which one fits is a question about the shape of your volume, not about your industry.
Minimum engagement is half a seat — not a five-seat block. We also bill per resolved issue, from $1, where you pay nothing if it is not resolved; that model fits businesses where a customer problem actually gets solved on the call, not businesses where the call is intake.
do the arithmetic yourself
Pull last quarter's port-outs and count, for each, how many support contacts that line generated in the ninety days before it left. Call that P. If P is above your average, support was not a cost centre on those accounts — it was the last place you could have intervened. Most carriers never run this, because support data and churn data live in different systems and nobody owns the join.
At a typical monthly ARPU, one retained line is worth more over its remaining contract than a year of that line's support contacts cost at $0.15 a minute. So the question is not how cheaply the queue can be answered — it is whether anything in it is being read as a signal. Repeat billing contact on one line, in a short window, is the most legible churn signal in this industry and among the cheapest to detect.
So split it: outage status, balances and plan lookups have one right answer and no retention risk — run them at the lowest rate that answers correctly, and spend what you saved on a human reading the queue for lines where P is climbing.
setup
Five days is our normal window to staff a live account — that is five days to find and place an agent who fits your business, not five days of software setup. AI voice can go live sooner because there is no hiring step.
your existing software
We work inside the CRM, practice system or case tool you already run, rather than handing you a second inbox to check. Every contact leaves a timestamped record — who called, what was asked, what was said back, what was promised — because in your line of work the question six months later is not "did someone call" but "what exactly did we tell them". Tell us what you run and we will say straight whether we connect to it today.
what comes off the call
Most of this already exists in your OSS or knowledge base. Send it up front because a telecom queue punishes vagueness faster than most: an agent who guesses at a proration rule creates a billing dispute.
The last one looks like a detail and is not. An account address and a subscriber's actual location stop matching the moment someone moves and keeps the number.
where AI should not be
"Why is my bill higher than last month?"
Almost always proration, and almost always correct — which is exactly why it needs a person. The customer is not asking for a calculation, they are asking whether they were treated fairly. An automated breakdown that is arithmetically perfect and tonally wrong turns a question into a dispute, and disputes here reach a regulator more often than in any other vertical on this site.
Anything to do with porting, in either direction
A port sits between two carriers and a customer who cannot be reached on the number in question. It is the one workflow where being stuck is the default rather than the exception, and it needs someone who can hold the case across days and chase the other side. A bot can report status. It cannot own a stuck port.
The third call this week about the same line
The first two may genuinely have been simple. The third is a person deciding whether this is worth the effort, and on a month-to-month plan that decision is a port-out request two weeks later. It needs someone with the authority to do something — a credit, a truck roll, a named callback — not a better explanation.
A callback that crosses a timezone
The constraint here is legal, not editorial: outbound to a residential subscriber is bounded by the hours at the called party's location, not the agent's.
a model we're not selling you
During a major outage, automation genuinely answers almost everything — and a vendor quoting deflection rate will show you a spectacular number for that week. It is the easiest number on this page to produce and the least informative. Nobody deserves credit for automating a question with one correct answer.
What the deflection figure hides is the residue. The calls left after the outage queue drains are the expensive ones — the customer whose business ran on that line, the one who has now called three times, the one asking why they are paying for a month they spent offline. A queue measured on deflection will show those as a rounding error in a very good week.
So we quote per-minute and per-session for the volume with one right answer, and a person for the volume that decides whether the line stays. Every model in our price list stays open — per resolution from $1 included, which fits billing work well. What we will not do is put a deflection target in a telecom contract: the week it looks best is the week you lost the most customers.
when we're the wrong choice
questions
AI voice is $0.15 per minute and web chat is $0.15 per session. Live agents start at $300 per month including 500 minutes — about $0.60 per minute — or from $5.80 per hour for a dedicated agent working only your account. The minimum engagement is half a seat, not a five-seat block.
Five days is our normal window to staff a live account — that is five days to find and place an agent who fits your business, not five days of software setup. AI voice can go live sooner because there is no hiring step.
Yes, on the same number, with agents in our Bolivia and Nicaragua operation centres — native Spanish, not machine translation. English and Spanish share one queue, and we cover 60+ more languages if you need them.
That is the normal setup. You decide which call types go to AI and which reach a person, and you can change the split as your volume changes. Nothing about that decision is locked in at signup.
The customer's clock, not your team's. Under 47 CFR 64.1200(c)(1) no telephone solicitation may reach a residential subscriber before 8 a.m. or after 9 p.m. local time at the called party's location — the called party's, explicitly. For a regional retailer that is one window. For a carrier with subscribers across four timezones it is four, and one agent can be compliant for part of a list and not the rest at the same moment. Ask any outbound vendor how the dialer enforces that per record, not per shift. If the answer is a start time, they are enforcing their own clock.
We can usually help you get one. We work through a network of dozens of carrier relationships rather than reselling a single supplier, and we can apply for numbers on your behalf across Europe, North America, the Middle East, Southeast Asia, Latin America, Japan and Korea. What we will not tell you is that we already hold numbers everywhere — provisioning depends on the jurisdiction, and some want a local legal entity and proof of address first. Name the country and we will tell you what it actually takes there.
It is the case per-minute and per-session pricing handles best, because you are not paying for the quiet weeks. What it does not handle is the staffing side of a major event — a fibre cut that triples volume for six hours needs people who already know your escalation path. Plan the shared-team layer for the baseline and agree the surge path before you need it, not during.
start
See how AI resolves subscriber issues faster than traditional call centers.