Shaping the Future of Voice UI

Your Voice moves the world.

We believe that "voices move the world," and with a hybrid team from Japan and Bangladesh, we founded the Voice AI platform Verbex. Humanity first created "voice," then invented text, printing, and the internet, driving the world forward.
However, text-centric communication widens the digital divide in countries with low literacy rates and causes stress even in countries with high literacy rates. If everything could be controlled simply by speaking, the world would be a much freer place. Verbex returns to its roots, enabling more relaxed and stress-free communication. The new UI is Voice. Your voice will move the world from now on. The Voice UI carries your voice and moves the world.

Our Mission

Connecting the world with voice

Our Vision

We will implement Voice UI in all use cases and redesign the world.

Our Strength

Development capabilities and business implementation capabilities for a unique voice dialogue AI model that "speaks like a human."

Our Journey

2017 Verbex 1.0 Release

We released our first prototype as an IVR-based solution. At the time, it only had basic speech recognition (STT), mainly focusing on recognizing numbers and simple words. Nevertheless, the "voice control" experience took shape, and this is where our Voice AI journey began.

Verbex 2.0 released in 2022

We have released a Rasa-based conversational engine for telephone lines.
Significant advancements have been made in both speech recognition (STT) and natural language processing (NLP), dramatically improving conversational context understanding and response accuracy. A total of 30 patents have been granted in Bangladesh, India, Indonesia, Nigeria, and Australia, greatly strengthening our technological foundation.

Verbex 3.0 released in 2024

We have released a new engine that utilizes generative AI to enable "human-like conversations."
Speech recognition (STT) now supports many dialects and noisy environments, and speech synthesis (TTS) achieves more natural and human-like speech.
Furthermore, we have released a top-of-the-line large-scale language model (LLM) for the Bangladeshi language.
Development for Japanese language support has also officially begun.

2025: Full-scale launch of business operations

Leveraging our accumulated technology, we are finally launching our business in earnest.
We will simultaneously launch services in multiple countries, including Bangladesh, Japan, India, and the Philippines. We will implement our unique voice dialogue AI, which we have cultivated so far, into society and spread the new possibilities of voice UI to the world.