Help and FAQ center

An AI voice call center uses an AI assistant to handle inbound and outbound phone conversations. AvaritCall lets you configure appointment, question-answering, sales and CRM logging workflows. Concurrent call capacity depends on the plan.

In AvaritCall’s published internal measurements, the average time from the caller finishing speaking to the agent’s first audible response is approximately 0.8 seconds in Türkiye. This is conversational response latency; call pickup time is a separate metric. Results vary with the connection, scenario and tool use. Measurement dates, sample sizes and p50/p95 distributions have not been published.

AvaritCall is optimized for Turkish phone conversations. Check availability for your target language and voice, and test your own call scenario before deployment.

Any business with phone traffic: healthcare and dental clinics (booking), real estate, automotive, car rental, e-commerce, education, finance, and customer support — from SMBs to enterprises that want to capture missed calls and stay reachable 24/7.

AvaritCall is pay-as-you-go, starting at $0.08 per talk minute with no commitment. Compared with a human agent staffing a 24/7 line, the per-minute cost is much lower — no shifts, training, or absenteeism. New sign-ups receive $5 in free credit.

No — no commitment, no contract. You start testing with a $5 welcome credit; to continue, you top up a minimum of $10 and pay only for the minutes you talk, at $0.08/min.

$0.08 per minute, including speech recognition, response generation and voice. Charged separately: email reports ($0.001), WhatsApp reports ($0.005), messages the assistant sends, IVR voice generation ($0.15/1K characters), $0.005 per call for every 10,000 characters of agent instructions, and 20% VAT added to top-ups.

Payments are processed securely via iyzico with a credit or debit card. You top up a minimum $10 balance and it's drawn down as you use it — no automatic card charges. You're notified in advance as your balance runs low.

Yes. The standard plan comes with high concurrent capacity; for very high volumes, the enterprise plan provides a dedicated server and custom capacity.

You can start in minutes with the ready dashboard: connect your number, add the assistant's knowledge base, and go live. API and webhook support is available for technical teams, but no coding is required.

Yes. AvaritCall doesn't run the call in isolation — it logs the conversation to your CRM, books into your calendar, and confirms over messaging. It connects to your workflow via API, webhooks, or ready-made flows.

You generate your API key from your dashboard right after signup. Connect your own systems via REST API and webhooks, and track calls, transcripts and minute usage live from the dashboard.

Yes, with RAG it answers from your PDF, website and text knowledge base.

Yes. No number porting is needed — a simple forward from your existing number to AvaritCall is enough. It runs alongside your PBX, and you can route only after-hours or overflow calls to the assistant if you prefer.

AvaritCall’s servers are hosted on AWS in the United States. Calls are not recorded as audio; only text transcripts are stored. Review the privacy and KVKK notices for data processing, providers and possible transfers. Clarify access, retention, deletion and any contractual requirements before use.

Yes. Your data stays encrypted, processed GDPR-compliant, and PII is auto-masked.

Calls are stored as text transcripts, not audio recordings, and data is kept encrypted on secure servers in line with KVKK (Turkish data protection law). On outbound calls the assistant never shares case details — only date, time and location. Your firm decides who can access the transcripts.

Yes. You upload a contact list, start a campaign, define the agent's script, and follow results from the panel and the iOS app. Common scenarios: appointment confirmation, payment reminders, lead qualification, win-back and satisfaction calls. Campaign results are reported per call with the transcript attached.

Yes, collections and payment reminders are among the most common uses. The agent states the amount and due date, secures a payment commitment and writes the outcome to your CRM. Because this is a sensitive area, you set the boundaries: what the agent may say, when it hands over to a human, and when it ends the call. Final financial decisions should stay with a person.

Yes; booking, confirmation and reminders are the most used scenarios in this sector, and evening or weekend calls stop going unanswered. No call audio is recorded — only text transcripts are stored, which reduces retention and consent burden on patient data. Diagnosis, treatment and medication guidance are outside the agent's scope and are handed to a human.

Two core differences. Packaging: those platforms bill speech recognition, the language model, voice and telephony separately; AvaritCall includes speech recognition, response generation and voice in one per-minute rate and charges no number or line fee (your line's call charges are with your carrier). Telephony: Retell and Vapi source it from third parties (Twilio, Telnyx), while AvaritCall builds its own telephony layer — codec handling, echo cancellation, connection and channel management in-house. Our comparison pages publish their current pricing from their own sources.

No. Audio is not recorded; only the text transcript of the conversation is stored. This is deliberate — it removes a large share of consent, retention and breach risk. If your QA process depends on listening back to calls, decide with that in mind: review is possible over transcripts, not over audio.

You define it. Common rules: the caller explicitly asks for a person, the topic falls outside the agent's knowledge scope, there is a signal of anger or complaint, or one of your defined trigger topics comes up. Conversation context is passed along with the handover, so the person does not start from zero.