Samantha AI, Romania’s first conversational artificial intelligence platform, has launched a new voice-based version that aims to deliver more natural and continuous user interactions. According to data quoted by the developer, users spend more time in voice interactions than in text-only sessions. 

Samantha AI was developed by Cristian Doiu and UXR, inspired by the operating system featured in the film Her. The platform is part of a broader digital ecosystem that includes Chatter.al, AWW.al, TheDesignBook.net, and Thinks.at.

The new speech-to-speech technology, developed locally for the first time, allows users to engage in uninterrupted voice dialogues. The platform includes automatic language detection with no prior settings required, as well as text-based interactions, text-to-voice, and voice-to-text capabilities.

“Samantha AI was designed from the very beginning as a social AI, not just a simple chatbot. Our goal was to build a natural, coherent, and easy-to-use conversational experience for real life. To continuously improve the experience, after each platform version, we analyse session length, user engagement, and preferences for voice or text. Samantha V3 shows that over 68% of users implicitly choose voice conversations,” said Cristian Doiu, founder of UXR. 

Samantha V1, based on text-only interactions, surpassed 20,000 visits. 

Meanwhile, Samantha V2, which introduced voice functionalities (voice-to-text), recorded over 100,000 visits in the first months, with an increase of over 400% in the volume of voice interactions. 

Samantha V3, currently in beta, indicates predominantly voice usage, with over 75% of total interactions conducted via speech-to-speech, its representatives said.

irina.marica@romania-insider.com

(Photo source: press release)