Enterprise AI CallcenterOn-premises
AI in call centres: statistics, facts and figures with sources
Every number on this page has a source or a measurement condition. We publish no estimates and no customer results without permission.
Our lab measurements (October 2026)
| Measurement | Value | Conditions |
|---|---|---|
| Simultaneous calls on one GPU server | 50,000+ | one of our enterprise GPU servers, standard setting; calibration with real speech |
| Voice first audio after the reply text is ready | 136 ms median, 178 ms p95 | same calibration, 50,000+ calls |
| Simultaneous calls, CPU only | 6 | the same server without its GPU |
| Inbound call answered | 0.14–0.31 s | SIP call through our telephony gateway |
| End of caller speech → first sound of the answer | 1.4 s median | 27 test calls in 14 days, cloud language model |
| Import of 100,000 contacts into a campaign | 20.6 s | same server |
External facts
| Fact | Source |
|---|---|
| EU AI Act, Art. 50: people must be informed that they are interacting with an AI system | eur-lex.europa.eu |
| GDPR, Art. 9: health data is a special category of personal data | eur-lex.europa.eu |
| Google FAQ rich results are limited to well-known government and health websites (since Aug 2023) | developers.google.com |
See it working on your own scenario
Book a 20-minute call, listen to the agent, or let us build a free sample agent for one of your call types. No commitment.
- GDPR by design
- Runs on your own servers
- 24/7 operation and support
- A team of 5 in Sofia, EU
Already running in production at organisations in Bulgaria — names under NDA.