Voice AI — inbound + outbound

It picks up the phone.Every call answered.Never rings empty.

Nortinia AI Call Center is a full voice-based customer service — inbound and outbound calls, wired into your ERP, unlimited languages. Picks up in 0.3 seconds, decides who to call back, books into the calendar, sends quotes, and can place parallel outbound calls to third parties while still talking to you. A 15-person call center runs €36-54k/mo. Nortinia AI Call Center Growth is €3,490/mo.

A 15-person call center runs €36-54k/mo. Nortinia AI Call Center Growth is €3,490/mo. 90%+ savings, 24/7 coverage, unlimited languages.

+47calls handled in the last hour
0.0s
Pick-up latency — the line never rings empty, even at lunch
0%
Cheaper than a 15-person call center
0
Languages — answers in whatever they call you in
0/7
Coverage — nights, weekends, holidays, vacations
15-person call center vs AI Call Center

Replace your call center with an AI.

An average 15-person inbound-outbound call center runs €36-54k/mo — salaries, payroll taxes, software licenses, telecom, training. Nortinia AI Call Center Growth runs €3,490/mo. Same work, at machine speed, in unlimited languages, 24/7, never quits and never goes on vacation.

Traditional setup

15-person call center + telecom + software

€36-54k/ mo
  • 8-16 hour shifts — nights, weekends, holidays go unanswered
  • One agent handles one call at a time
  • 3-6 months of onboarding for new hires
  • Native local language + broken English — anything else is impossible
  • 40%+ annual turnover — constant recruitment + training
  • Telecom + software licenses + CRM extra: €2-3k / mo

Plus the CEO's time scheduling shifts. Plus the lost customer who calls at 7pm Friday and no one picks up. Plus the 4-6 minute average hold time that makes them hang up first.

Nortinia AI Call Center

One AI voice customer service

€3,490/ mo
  • Picks up in 0.3s — line never rings empty, no one waits in queue
  • Handles 20-100 calls in parallel, depending on plan
  • Live in 4-8 weeks — then it knows your business processes
  • Natively in unlimited languages — EN, HU, DE, ES, ZH, anything
  • Never quits, never goes on holiday, never errors during onboarding
  • ERP + CRM + calendar + email in one thread — every call documented

Bigger one-time investment — you have to map out the company's call workflows and communication scripts. After that it runs 24/7, in unlimited languages.

What companies use it for

12 scenarios. All from real customers.

AI Call Center isn't an IVR tree, and it's not a chatbot in voice-only mode. It's a full-featured phone agent that knows your business processes and can execute ERP-level transactions. Below are 12 concrete scenarios — exactly how customers use it.

Tourism · parallel call

Live negotiation with accommodation mid-call

A visitor books accommodation through a travel agency, but has a special request not on the website. They call the agency. While they're talking to the AI Call Center, the AI in parallel calls the accommodation, sorts the issue out in the right language, and returns to the caller with the answer — without a single pause in the conversation.

Legal · instant knowledge synthesis

Legal consultation in seconds

"I need legal info." The AI scans your entire internal policy + relevant statutes in seconds, then delivers expert-level advice. Depending on your delegation level, it can even represent the company — drawing from whatever documents it has access to.

B2B sales · outbound campaign

200 cold calls in parallel

On an outbound sales campaign the AI dials 200 leads at once, voice-led but with all 14 channels pre-warmed. It qualifies intent and only routes hot leads into the salesperson's actual calendar — booked, not promised.

Healthcare · appointment booking

Patient calls, AI books the slot

The patient calls the clinic. The AI checks the doctor's calendar, books the right slot, confirms via SMS, and files the visit into the EHR. With consent, it can also collect pre-visit symptom info after security verification.

E-commerce · order status

"Where's my order #12345?"

The customer calls support with their order number. The AI runs 2-factor authentication (e.g. date of birth + last 4 of card), then pulls live status from the ERP, gives the ETA, and SMS-es the tracking link.

Multilingual · automatic switch

Starts in Hungarian, continues in English

The caller starts in Hungarian, then mid-call mentions they're abroad. Within one sentence the AI switches to English (or German, Spanish, Chinese — whatever they're really speaking) and continues in their native language. The caller never feels the transition — just a natural language switch.

Insurance · claim intake

Intake + parallel garage check

The caller reports an incident. The AI gathers structured data (date, location, parties involved, damage), then in parallel dials 3 partner garages, checks their capacity, and returns to the caller with the 3 nearest available slots — in a single conversation.

Real estate · 3 viewings

3 owners' calendars at once

The prospect wants to see 3 apartments. The AI in parallel checks 3 owners' calendars, finds the 2 time slots where all three are available, and books the listing agent in too. One call, three coordinated bookings.

HR · internal ticket triage

Employee HR line

An employee calls the internal HR line. The AI infers intent (payroll vs benefits vs compensation), and either routes to the right HR officer, or — if it's a simple document request (e.g. employment letter) — handles it itself after identity verification.

After-hours · emergency

2am incident report

A customer calls outside hours. The AI gathers the issue, classifies urgency, and for critical cases pages the on-call team (SMS + phone). For non-critical cases it files a ticket and promises a callback by morning — automatically scheduled.

Membership · renewal

Annual renewal call

The AI calls existing customers 30 days before renewal, offers loyalty pricing, and closes the renewal live on-call — confirming payment, emailing the new contract. It can also reschedule the call if the customer isn't free right now.

PR · press inquiry

Journalist on the press line

A journalist calls the press line. The AI provides the press kit, verifies embargo status, and if an exclusive interview is requested, escalates to the communications director — while logging the journalist's name, outlet, and topic.

The killer feature

Parallel call handling — talks to you while dialing the next party.

This is the capability that makes Nortinia AI Call Center the highest-tier product in the stack. While you're talking to it, the AI can in the background start another outbound call — to a partner, a hotel, a supplier, a colleague — handle the matter in the right language there, and come back to your conversation with the answer. The flow never breaks. A human agent literally can't do this. An AI can.

1

You call in — the AI picks up in 0.3s

The line never rings empty. The AI picks up instantly and listens to your need — in any language you call in. It recognizes returning callers.

"No queue. No IVR tree."

2

Needs info from a third party — AI dials in parallel

You mention you have a special request for the hotel. The AI <em>mid-conversation</em> calls the hotel in the right language, while keeping your conversation going. You stay on the line, chatting through the details.

"Works in the background. You don't feel a pause."

3

Brings the answer back, closes the call

The AI returns with the answer. "I confirmed with the hotel — they've held a dinner slot, and the allergen-free menu is available." It books into the calendar if needed and sends an email confirmation.

"One call. One issue. Six transactions executed."

What it does, exactly

Twelve capabilities that separate a voice AI from an IVR tree.

Not a feature list — a results list. Each item is concrete work you used to pay people for, or had no capacity to do at all. Real, voice-based, ERP-wired phone handling.

01

Inbound + outbound

Works both directions. Inbound customer service (support, booking, order status) and outbound campaigns (sales, renewals, callbacks) — same AI, same knowledge base.

Both ways
02

Parallel call handling

From the Growth tier: while talking to you, the AI places parallel outbound calls to partners, suppliers, colleagues — and weaves the answer back into your conversation. The killer feature.

Growth+
03

Unlimited languages, natively

On Scale, languages are unlimited. The AI answers in the caller's native language, accent-aware. Hungarian formal/informal, German du/Sie, Spanish tú/usted — all on point.

Global
04

Deep ERP + CRM integration

Calendar booking, invoice status, stock query, quote sending, order entry — all through your business systems, in real time. HubSpot, Salesforce, Pipedrive, custom webhooks.

Live transactions
05

Security verification

Before releasing sensitive data, the AI verifies the caller — date of birth, last 4 of card, unique PIN, voice biometric, or SMS-OTP combinations. Never gives data to an unverified caller.

GDPR + PSD2
06

Soft handoff to human agents

If the AI judges it's beyond scope, or the caller explicitly asks, smooth handoff to the right colleague — with a pre-brief so the customer doesn't repeat themselves.

Soft handoff
07

Call recording + transcript + AI summary

Every call recorded (consent-based, opt-in), automatic transcript, and a 3-5 sentence AI summary into your CRM at the end of every call. Audit, training, dispute resolution all built on this.

Compliance
08

Sentiment + escalation routing

Detects when someone's frustrated, angry, or here to buy. Decides whether to route to sales, support, or retention — or whether it can close the matter itself.

Realtime
09

Phone number provisioning

The platform provisions phone numbers per country, brand, campaign. White-label, local, mobile — backed by Telnyx + alternate carriers.

Per-region
10

Quote + booking authority

Where authorized, the AI sends live email quotes during the call, books into the calendar, or issues invoices — all according to company rules (discount limits, approval flows, escalation).

Transactional
11

Voice cloning (Growth+)

Custom voice profile — we record 15-30 minutes from your brand's chosen voice, and the AI takes on that character. Consistent acoustic brand on every call.

Brand voice
12

Telnyx-grade reliability + failover

Carrier-grade infrastructure, automatic failover if a number / carrier goes down. European data centers, GDPR-compliant storage, 99.9% SLA on Scale.

Carrier-grade
Not an IVR, not a chatbot — voice AI

Forget the "press 1 for English" decision-tree bots.

People hate IVRs because they're static rulebooks that hold you for 4-6 minutes then route you to another rulebook that holds you again. Nortinia AI Call Center is a real voice agent: it understands natural speech, knows your processes, works with your ERP, and executes transactions — exactly the work you used to pay humans for.

  • No "press 1" — speak naturally, say what you want
  • Not a rulebook — contextual reasoning, ERP queries, real transactions
  • Unlimited languages — no separate call center per region
  • Parallel call handling — a single human agent literally can't do this
  • Audit trail on everything — recorded, transcribed, AI-summarized
  • Soft handoff to a human, with a pre-brief

Typical daily volume at a customer

Calls answered1,247
Parallel outbound (avg)84
Human handoffs29
Avg pick-up latency0.3s
Languages in use11 / unlimited
Customer CSAT4.8 / 5
What you're replacing

How to calculate? — 3 scenarios.

This visual ROI helper helps you decide which tier to pick. Nortinia AI Call Center bills per-second (displayed per-minute), and every tier includes prepaid minutes. Overage minute rate drops by tier — Scale is one-third of Starter's per-minute price.

Small-medium ltd (3-person receptive CC)~500 min/mo · 1 language · standard ERP integration
Starter
€1,490 fixed · 500 min included
Growing scaleup (10-person CS team)~1,500 min/mo · 7 languages · parallel calls · CRM integration
Growth
€3,490 fixed · 1,500 min included (+~€500 per-min)
Enterprise (50+ agent equivalent)~8,000 min/mo · unlimited languages · multi-tenant · 24/7 SLA
Scale
€7,990 fixed · 10,000 min included · overage €0.06/min
Custom / on-premDedicated VPC, compliance pack, dedicated infrastructure
Enterprise
Custom quote
Average 15-person call centerSalaries + payroll taxes + software + telecom
Traditional
≈ €36-54k / mo
Nortinia AI Call Center Growth replaces itSame volume, 2x scalability, 24/7 coverage
AI
€3,490 / mo
Typical monthly cost of a 15-person call center€36-54k
Nortinia AI Call Center Growth€3,49090%+ savings · 24/7 · unlimited languages

A 15-person call center doesn't answer at 10pm on a Friday. Your customer calls at 10pm on a Friday.

Minutes + tokens, transparently

Per-second billing. Token allowance. No surprises.

Every Nortinia AI Call Center plan includes prepaid minutes AND an LLM token allowance. Minutes cover the voice/telecom layer, tokens cover the conversational intelligence. Overage flips automatically and gets a per-tier discount. New standard across all Nortinia products.

Starter
500 min + 3M tokens
≈ 500 typical calls (~1 min / call)

Overage rate€0.15 / min · €0.025 / 1k tokensStarter rate
200 extra minutes€30
Growth
2,000 min + 40M tokens
≈ 1,500-2,500 calls / mo

Overage rate€0.10 / min · €0.018 / 1k tokens33% per-min discount
1,000 extra minutes€100
Scale
10,000 min + 200M tokens
≈ 8,000-12,000 calls / mo

Overage rate€0.06 / min · €0.012 / 1k tokens60% per-min discount
5,000 extra minutes€300

What's "one minute"? The platform bills per-second internally, displayed rounded to the minute in the UI. A typical order-status lookup runs 45-90 seconds, an inbound booking call 2-4 minutes, an outbound sales call 3-7 minutes. The AI Engine optimizes the token allowance: fast lookups on a cheap model, complex negotiation on the stronger model.

How to start

Process discovery + integration + go-live. Bigger investment — then 24/7 it works.

Honest version: this is not a "paste a script tag and go" deploy. Rolling out the Nortinia AI Call Center is a real company project — you have to map every phone workflow, communication script, ERP integration, escalation rule. The consulting package isn't optional — it's required. That's why we recommend the Starter consulting + Call Center onboarding combo. Without it, deployment will struggle and likely fail.

  1. 1

    Process discovery (Starter consulting)

    We walk through your entire phone workflow: who calls, with what, where the info goes, who decides, who quotes, how things escalate. Workshop format, 3-5 sessions. Output: documented call scripts.

    2-3 weeks
  2. 2

    ERP + CRM + calendar integration

    We wire the AI into your business systems — HubSpot, Salesforce, Pipedrive, or custom ERP webhooks. Calendar, invoicing, inventory, customer database. Security layer + audit log on everything.

    2-4 weeks
  3. 3

    Voice training + go-live + monitoring

    Brand voice capture (15-30 min sample), call script fine-tuning, A/B test on 100-200 real calls. Phased go-live: 10% traffic first week, 100% by week three. Continuous monitoring + weekly business review.

    3-5 weeks

Our consulting team carries you through the whole project. We DO NOT recommend deploying the Call Center without consulting — too many traps (workflow gaps, ERP integration pitfalls, legal-compliance questions). Nortinia Consulting offers Starter / Growth packages dedicated to this — pick one alongside your Call Center onboarding.

Generational shift

If your competitor switches to AI Call Center and you don't — your customers are calling their line, not yours.

The previous generational shift in phone customer service was IVR + outsourced call centers — around 2005. Whoever missed that is still putting customers on 8-minute hold. The next shift is AI Call Center. It's happening now. Whoever starts 12-18 months from now will be behind — and the customer will already have gotten used to the competitor's fast, multilingual, 24/7-available voice service.

— Norbert Kubinyi, founder

Your customer already expects to be picked up in 0.3s.

A Gen Z buyer is not going to wait 4 minutes at the end of an IVR tree. They'll hang up and try the next result. AI Call Center picks up in 0.3 seconds and speaks like a human.

"Office hours" is dead.

38% of inbound leads call after 7pm, on weekends, or on holidays. You're asleep or on vacation. Your competitor's AI Call Center is closing the deal.

"Please call back later" will be unacceptable by 2027.

AI Call Center never says that. It picks up every time, always has an answer or can execute a transaction. Whoever still runs a manual call center in 2 years will be paying 5x more than the market — and going slower.

Separate inbound call center + separate outbound sales + separate language lines = €36-54k/mo. Miss this generational shift → 24/7 your competitor's line picks up instead of yours.

Simple pricing — voice AI

One AI call center's monthly price. A 15-person team's results.

EUR-denominated, billed monthly. 20% off with annual commit. The monthly fee includes the platform AND the included minutes + token allowance AND the Telnyx call infrastructure. Onboarding (process discovery, ERP integration, voice training) is a separate one-time project with the Nortinia consulting team — we don't recommend going live without it.

Starter

The first call center that works at night.

€1490/ mo
  • 3 concurrent call lines
  • 500 minutes + 3M tokens included
  • HU + EN, natively
  • Standard ERP integration (HubSpot/Salesforce/Pipedrive)
  • Call recording + transcript + AI summary
  • Email + ticket-handoff support
  • Inbound + basic outbound
Let's start

500 min / mo · overage €0.15/min · 3M tokens · overage €0.025/1k

Scale

Global call center at machine speed — fully AI-driven.

€7990/ mo
  • 100 concurrent call lines
  • 10,000 minutes + 200M tokens included
  • Unlimited languages
  • White-label phone numbers (per-region provisioning)
  • Multi-tenant routing (multi-brand)
  • 24/7 SLA, dedicated customer success
  • Compliance pack (PCI-DSS, HIPAA-equivalent)
Let's talk

10,000 min · overage €0.06/min (-60%) · 200M tokens · overage €0.012/1k

Enterprise

When the company size demands it.

Custom
  • Unlimited concurrent lines
  • On-premise / dedicated VPC
  • Custom Telnyx routing + your own carrier
  • Compliance pack (PCI-DSS, HIPAA)
  • Custom voice training
  • Dedicated infrastructure + 99.99% SLA
  • Dedicated customer success team
Get a quote

On request — custom quote

Stronger together

Sales AI brings the leads. Call Center handles, qualifies and closes the calls.

Pair Sales AI with the Call Center and the lead-flywheel closes itself. Sales AI generates outbound leads across 14 channels and produces content + PR signal; AI Call Center picks up the resulting inbound calls and runs the outbound dial-out — at machine speed, in parallel, with full ERP integration.

Upstream

Nortinia Sales AI

  • Outbound lead generation across 14 channels
  • Content + thought leadership pipeline
  • PR + outreach automation
  • Cold email + LinkedIn DM + WhatsApp orchestration
  • Lead scoring + CRM hand-off
Open Sales AI
Downstream — this site

Nortinia AI Call Center

  • Voice intake of every inbound lead, 0.3s pick-up
  • Parallel outbound dial-out from the Sales AI queue
  • Hot-lead booking straight into the salesperson's calendar
  • Recording + transcript + AI summary into the CRM
  • Multilingual coverage across every market

Pricing tip: Sales AI Growth €1,990 + Call Center Growth €3,490 = €5,480 / mo — fully replaces the marketing-and-call-center stack of a 25-person revenue team.

Two products, one ecosystem

AI Call Center vs AI Assistant — honestly.

Both products run on the same Nortinia AI Engine. The difference: Call Center handles unlimited parallel voice calls with recording, transcripts and transfers. The Assistant runs task-specific back-office automations (invoice matching, procurement, reporting) — it's NOT a call center. They work best together: Call Center picks up the phone, Assistant does the back-office work.

CapabilityAI Call CenterAI Assistant
Voice inbound call pickupYes — 0.3sNo (back-office)
Voice outbound campaignsYes — in parallelNo
Parallel call handlingYes (killer feature)N/A
Call recording + transcript + summaryYesNo voice
Phone number provisioningYesNo
Voice cloning / brand voiceYesNo
Back-office workflows (invoice, reports)Only via voice handoffNative strength
Scheduled task automationLimitedNative strength
Unlimited languagesYes (Scale)Yes
ERP + CRM integrationYesYes
Growth tier monthly fee€3,490€899
FAQ

What every CEO asks first.

No marketing fluff. Concrete numbers, deadlines, answers.

How much to get started?
Two line items: (1) the monthly platform fee (Growth €3,490 — includes 2,000 min + 40M tokens / mo) — paid monthly even during onboarding; (2) a one-time onboarding project with the Nortinia consulting team — process discovery, ERP integration, voice training, 7-12 weeks. Project pricing depends on company complexity — a simple 5-person business starts at €1.5k, an enterprise customer from €15-45k. One-time.
How long until it's live?
Bigger investment than chat — 7-12 weeks typically. A simple small business (1 language, 1 ERP, 1 call script) is 6-8 weeks. A complex enterprise (multi-brand, multi-language, multi-ERP, compliance) is 12-16 weeks. The "flip it on in 5 minutes" pitch is straight-up lying for voice AI — your scripts, ERP integrations, and escalation rules need to be mapped, and that takes time.
Why is the consulting package mandatory?
Nortinia AI Call Center executes real transactions against your business systems and speaks for your company on the phone. If processes, escalation rules, legal boundaries, and voice samples aren't right, the AI will make bad decisions — and that's expensive. The consulting team prevents that: maps the workflows, writes the scripts, sets the fail-safes. <strong>That's why we DO NOT recommend the Call Center without consulting.</strong> The combo: Starter consulting + Call Center Starter package.
Does it really place parallel calls?
Yes — this is the killer feature. On Growth and Scale: while you're on the line with it, the AI can place outbound calls in the background — to partners, hotels, suppliers, colleagues — handle the matter in the right language, and return the answer to your conversation. Your call never breaks. A human agent literally cannot do this; an AI can.
What happens if I blow past the minute or token allowance?
It doesn't stop. Auto-flips to overage billing: Starter €0.15/min, Growth €0.10/min (-33%), Scale €0.06/min (-60%). Same flow for tokens. A typical order-status lookup runs 45-90s, an inbound booking 2-4 min. The monthly report itemizes spend, and we send an 80% utilization warning.
GDPR-OK?
Yes. Every call is logged server-side (opt-in basis with proper notice — "this call may be recorded for quality"), PII-masked. Data stays in the EU. DSAR + erasure API built in. The caller can request or delete their data anytime. PCI-DSS and HIPAA-equivalent compliance pack on Scale.
What languages?
Starter: HU + EN. Growth: HU + EN + DE + ES + IT + FR + ZH (7 main languages). Scale and Enterprise: unlimited, added on demand. The AI answers in the caller's native language, accent-aware — HU formal/informal, DE du/Sie, ES tú/usted — all on point.
How does it release sensitive data?
Never without security verification. Verification methods: date of birth, last 4 of card, unique PIN, voice biometric (Growth+), SMS-OTP, emailed magic link — or combinations, per your company's rules. If verification fails, automatic escalation to a human colleague.
Can I switch to the Assistant later if I don't need voice?
Yes. Both products use the same AI Engine — knowledge base, brand voice, CRM integration are all portable. Drop the voice module, keep the Assistant. Or the reverse: add a Call Center module to an existing Assistant deployment.
What if I cancel?
Every piece of data (call history, transcripts, AI summaries, lead data, knowledge base export) is exportable with one click, JSON or CSV. Numbers shut off immediately. No cancellation fees. You own the data, always.
Does it really replace a 15-person call center?
The Growth tier (20 parallel lines) covers an average 15-20 person call center's traffic. The Scale tier (100 parallel lines) covers a 50-80 person enterprise CC. ARPC (average revenue per call) is 30-50% higher because parallel call handling closes more transactions within a single customer conversation. Typical numbers: 70-80% of calls fully closed by AI, 20-30% selective human handoff — and that 20-30% is usually genuinely human-decision work.
What does voice sampling mean?
From Growth: we record 15-30 minutes of audio from the brand's chosen voice (CEO, customer service lead, or a hired professional voice talent). The AI learns the voice's character, accent, cadence, and uses it for every call — consistent acoustic brand. Legal consent is mandatory (treated as GDPR biometric-style processing).
Let's talk

Bring us a workflow map. Two months later it answers your calls.

We reply within one business day. Process consultation + demo call + quote. In Hungarian and English.

Response timeWithin 1 business day