Aculab Cloud vs LumenVox Active Voice Biometrics

Aculab Cloud provides APIs for building applications that handle voice and video calls, messaging, and inbound or outbound fax. It supports calls to telephone numbers and SIP addresses, and its WebRTC interface can place live calling in a web page. Listed capabilities include live speaker detection, speech recognition, translation, text-to-speech, voice biometrics, SMS, and MMS. The REST API works with Python, C#, VB, Java, PHP, or other scripting languages; REST applications run on a web server you host. The UAS API supports Python, C#, VB, F#, and C++, with applications running in a User Application Server installed on your platform. Speech recognition uses Google Speech-to-Text, while text-to-speech supports Amazon Polly and Cepstral engines. Aculab Cloud runs on AWS; web services use HTTPS, UAS connections use SSL by default, and customers can encrypt media files with keys supplied at runtime. Accounts are assigned to Europe, USA-East, or USA-West (beta). The free Trial includes $10 in credit for one month. Production is pay-as-you-go, with rates shown in the Cloud Console; Premium usage is invoiced and requires contacting Aculab.

LumenVox Active Voice Biometrics provides text-dependent voice authentication for enterprises seeking identity assurance beyond passwords, PINs, and security questions. Containerized microservices run on Kubernetes on-premises, in private or public clouds, or across hybrid environments. A deep-neural-network engine supports active, passive, and hybrid authentication. Enrollment and verification can use any passphrase and language supported by the corresponding ASR engine. Playback, liveness, synthetic-speech, and fraud-watchlist checks address presentation attacks. APIs create, retrieve, disable, re-enable, and remove identities and query enrollment status. Biometrics, Speech, Management, and Reporting APIs are exposed through gRPC and REST. Supported channels include contact-center IVR, smartphone and web applications, desktop applications, virtual assistants, video calls, messaging platforms, and conversational AI. Enrollment requires at least three separate utterances. Audio must be headerless base64 PCM-16, A-law, or μ-law at 8 kHz. Data is encrypted at rest and in transit, with customer-controlled keys and TLS between components. Pricing is on request.