A Voice That Responds in Under a Second
At the core of the product is voice interaction that keeps conversations flowing with sub-second responses. It detects and switches between more than 90 languages on the fly, and voices can be selected from over 10,000 options, with voice cloning also supported. Responses that account for emotion and context, along with natural turn-taking, are cited as design goals. More than 5 million agents are in operation, and the company says conversations can scale instantly to millions of concurrent users.
Bridging the Phone Network and Business Operations
A distinguishing feature is its connection to the telephone network. It links, in many cases with a single click, to Twilio, Genesys, Vonage, Telnyx, Plivo, and SIP-capable PBX systems. On the business side, documents and FAQs can be ingested to ground responses, and handling procedures and escalation rules can be defined without code. Use cases cited include telecom support, financial inquiries and identity verification, medical reception and triage, and retail product guidance, with Revolut, Klarna, Deutsche Telekom, and Epic Games listed among its customers.
How It Differs from Chatbots
Conventional chatbots rely on rules and keyword matching and struggle with unexpected questions. Conversational AI, by contrast, understands intent and context, handles ambiguity, maintains memory across multiple exchanges, and can compose responses to questions it has never encountered before — so the difference is explained on the product side.
Security and Constraints
Communications are encrypted both in transit and at rest, with compliance with SOC 2, HIPAA, and GDPR, and options for data residency and zero data retention, according to the company. Pricing is not listed on the public page; only an inquiry contact is provided. While a reception completed entirely by voice can now be built at low cost, managing whose voice speaks to whom is left to those deploying it.
この記事を閉じる3つの問い
- When a voice receptionist can be built without code, what should front-desk staff turn their attention to instead?
- When identity verification and medical intake are handled by voice alone, who bears responsibility for misidentification and impersonation?
- A receptionist that speaks 90 languages — what will it flatten among the markets once carved up by the language barrier?