3i the VOX Platform for processing and speech analysis

Main articles:

• Speech technologies: on the way from recognition to understanding
• Technologies of biometric identification

3i VOX is the cloud AI platform for processing and speech analysis through API. The platform allows to implement functionality of speech recognition, voice identification (biometrics) and understanding of the natural speech in any third-party applications and products by means of HTTP requests to the relevant services through uniform API. A product it is focused on system integrators, software developers, vendors and corporate IT structures.

3i VOX can be used for any tasks where it is required to distinguish, understand and identify the speech. Based on the platform it is possible to implement different solutions, products and services, such as: speech analytics, voice bots, antifraud, DLP, etc.

The platform is completely built on own developments entered in the register of domestic software of the Ministry of Telecom and Mass Communications

2021: Development of models for speech recognition in the Kazakh and Uzbek languages

Company 3iTech 15 of January, 2021 announced development of speech models for speech recognition in the Kazakh and Uzbek languages, in addition to Russian and English languages, opened access to their testing. These languages will expand possibilities of platform 3i VOX as ecosystems of speech services for developers, integrators and corporate IT departments which use speech technologies.

Updating for January, 2021 is already available to testing in a personal account and through API. These language models will be interesting to the companies and developers who use speech technologies in creation of intellectual products and solutions for the markets of the CIS.

The Central Asian markets – one of development priorities for company 3iTech. Therefore we began to support by the first of the Russian developers Uzbek, having estimated the market potential and demand of speech technologies for local business and the state customers. Unlike Uzbekistan at Kazakhstan there are both Russian developers, and global players. However and there is a pent-up demand for the speech technologies and services constructed on their basis. Creating speech models of the Uzbek and Kazakh languages, we closely interacted with our regional partners. In 2021 our platform will be able already to be used as the domestic companies working for these markets and local players, - Alexey Lyubimov, the chairman of the board of directors of company 3iTech says.

Services of speech recognition (ASR), speech synthesis (TTS) and identification on a voice (Voice ID) are available to users of platform 3i VOX. Services 3i VOX through uniform API are built in products and solutions of the third-party companies and can be used in the systems of speech analytics, the systems of a voice antifraud, DLP systems, voice bots and also in services of subtitling of direct TV-air and podcasts. Deployment of services in an IT circuit (on-premise) of the customer is also possible. The platform is focused on software developers, system integrators, IT department of the companies and startups. AI platform 3i VOX is built on own developments of company 3iTech which are entered in the Register of domestic software of the Ministry of Telecom and Mass Communications.

2020

System of speech synthesis 3i TTS

Company 3iTech expanded tool kit for developers, having created the system of speech synthesis 3i TTS. A system became a part of an ecosystem 3i to VOX which is applied as for creation of difficult corporate solutions, and web services. 3i TTS can be used during creation of the systems of robotic call-down, voice assistants and intellectual IVR, reported on December 3, 2020 in 3iTech.

The system of speech synthesis 3i is intended to TTS (Text-To-Speech, TTS) for voice assistants and intellectual IVR and also scoring of text materials, development of automatic systems of voice notifications and notifications. Ecosystem products 3i VOX is the speech recognition engine (3i ASR), the system of speech synthesis (3i TTS), voice biometrics, the system of a diarization, etc. – developers, system integrators and solution providers for contact centers use.

3i TTS uses the NLP processor and neural network text normalization. A system works for CPU in real time. It is available through API or it can be unrolled on the customer's IT infrastructure.

"Speech synthesis became the last brick when forming our technology ecosystem. We offer partners a full range of tools for work with the speech within a uniform ecosystem 3i VOX. We have everything that developers could build the systems of any level of complexity. And, we suggest customers the choice – to use tools through API or to unroll the on-premise system in their IT circuit", - Alexey Lyubimov, the chairman of the board of directors 3iTech noted.

An opportunity to understand a youth slang and confused informal conversation

Company 3iTech 22 of May, 2020 announced that it optimized speech model which is used in the system of speech recognition 3i by ASR. Now platforms which are built on this speech engine (the platform of speech analytics 3i TouchPoint Analytics and cloud AI platform 3i VOX) will be able "to understand" a youth slang and confused informal conversation. Read more here.

3iTech submitted AI platform 3i VOX

Company 3iTech in April, 2020 opened access to qualitatively new version of the cloud AI platform for processing and speech analysis 3i VOX. Upgrade affected all basic elements of the platform: the speech recognition engine 3i is integrated by ASR 2.0, service of voice identification is built-in and the personal account for management of API keys and services is updated. Anyone can test quality of recognition on own records via the web interface.

Services of speech recognition (ASR), natural languag processing (NLP/NLU) and identification on a voice (Voice ID) are available to users of the platform. Services 3i VOX through uniform API are built in products and solutions of the third-party companies and can be used in the systems of speech analytics, the systems of a voice antifraud, DLP systems, voice bots and also in services of subtitling of direct TV-air and podcasts. The platform is focused on software developers and intellectual services, system integrators, IT department of the companies and technological startups. In new version 3i VOX there is a simple and functional personal account which significantly facilitates interaction with service.

The platform is built on own developments of company 3iTech which are entered in the Register of domestic software created the Ministry of Telecom and Mass Communications. ASR 2.0 engine 3i which is used by the platform increases efficiency of computing means more than for 30%, at the same time providing for 15% higher accuracy of recognition, than the previous version. The engine allows to distinguish correctly even those words which are not in the dictionary. At the same time, the additional possibility of improvement of quality at the expense of the intelligent mechanism of adaptation of a language model under data domain and specialized lexicon is included.

"Upgraded version of a cloud platform is, in fact, absolutely new product which uses more powerful speech engine. New interfaces are developed for users of the platform. Of course, 3i VOX, is focused first of all as before on developers who use speech technologies in the products. However, now also others can use it market participant. Start of the new version of the platform makes more available building tools therefore promotes growth of the market of intellectual business applications and services", - Alexey Lyubimov, the chairman of the board of directors 3iTech says.

Company 3iTech provided the test period for acquaintance with opportunities of AI platform 3i VOX.