Solutions

Using state-of-the-art speech recognition to provide customised solutions for under-resourced languages

Using state-of-the-art speech recognition to provide customised solutions for under-resourced languages

State-of-the-art Transcription

Transcribing under-resourced languages

Media Monitoring

Broadcast media monitoring with accurate customised keyword detection

Speech Analytics

Tailored for call centre dialogues

Speech Recognition for
Broadcast Media Monitoring

Speech Recognition for Broadcast Media Monitoring

Broadcast media monitoring is crucial for organisations to manage their reputation by staying informed about brand portrayal. It provides valuable market insights, competitor analysis, and public sentiment data, informing strategic decisions. Additionally, it ensures compliance with regulatory standards and measures the effectiveness of media campaigns, helping optimise future efforts.

Our solution allows media monitors to precisely track brand coverage across broadcast media. 

  • Highly accurate results

  • Unique keyword recognition

  • Code-switching for multi-lingual conversations

Broadcast media monitoring is crucial for organisations to manage their reputation by staying informed about brand portrayal. It provides valuable market insights, competitor analysis, and public sentiment data, informing strategic decisions. Additionally, it ensures compliance with regulatory standards and measures the effectiveness of media campaigns, helping optimise future efforts.

Our solution allows media monitors to precisely track brand coverage across broadcast media. 

  • Highly accurate results

  • Unique keyword recognition

  • Code-switching for multi-lingual conversations

Callbi speech analytics
for call centres

Callbi speech analytics for call centres

Easily import and search call recordings for insights to improve business performance. Callbi Speech Analytics utilises our advanced technology to provide cloud-based, software-as-a-service speech analytics for call centres.

  • The top speech analytics solution in SA

  • Multilingual: transcribes English, isiZulu, Afrikaans, Sesotho and Setswana

  • Optimised for call centre dialogues

  • Cloud-based

Easily import and search call recordings for insights to improve business performance. Callbi Speech Analytics utilises our advanced technology to provide cloud-based, software-as-a-service speech analytics for call centres.

  • The top speech analytics solution in SA

  • Multilingual: transcribes English, isiZulu, Afrikaans, Sesotho and Setswana

  • Optimised for call centre dialogues

  • Cloud-based

State-of-the-art Transcription

State-of-the-art Transcription

This cutting-edge ASR system specialises in transcribing under-resourced languages, providing speakers of these languages with access to technology that has historically been available only to speakers of a few major languages.

  • Cost effective

  • Multilingual transcription

  • SA English, isiZulu, Sesotho, Afrikaans

This cutting-edge ASR system specialises in transcribing under-resourced languages, providing speakers of these languages with access to technology that has historically been available only to speakers of a few major languages.

  • Cost effective

  • Multilingual transcription

  • SA English, isiZulu, Sesotho, Afrikaans

Real-time Transcription

Real-time Transcription

Real-time automatic speech recognition (ASR) instantly converts speech into text as it is spoken. Saigen offers accurate real-time speech-to-text models trained on conversational data that can be fine-tuned with your own speech data for even greater accuracy.

Our multilingual real-time ASR service provides live partial results within milliseconds, while adapting these partial results further in the background for more accurate “final” transcription results. Our multilingual ASR models can transcribe multiple South African languages in a single utterance or conversation.

  • Initial results in milliseconds

  • Trained on conversational data

  • Even greater accuracy when fine-tuning with your data

  • Enables real-time ASR for chatbots and conversational AI

  • Supports English, isiZulu, Sesotho and Afrikaans

  • Highly scalable

Real-time automatic speech recognition (ASR) instantly converts speech into text as it is spoken. Saigen offers accurate real-time speech-to-text models trained on conversational data that can be fine-tuned with your own speech data for even greater accuracy.

Our multilingual real-time ASR service provides live partial results within milliseconds, while adapting these partial results further in the background for more accurate “final” transcription results. Our multilingual ASR models can transcribe multiple South African languages in a single utterance or conversation.

  • Instant initial results in milliseconds

  • Trained on conversational data

  • Even greater accuracy when fine-tuning with your data

  • Supports English, isiZulu, Sesotho and Afrikaans

  • Highly scalable

Text to Speech (TTS)

Text to Speech (TTS)

Saigen’s Text-to-Speech (TTS) technology converts text into high-quality, natural-sounding synthetic speech in real time. Our multilingual models currently support English and Afrikaans, with isiZulu, Setswana, and Sesotho in active development.

Key Features:

  • Custom voice models tailored to your brand’s tone, pace, and pronunciation

  • Natural pacing and pronunciation for improved user engagement

  • Multilingual support: English, Afrikaans, and more South African languages coming soon

  • Scalable B2B-ready APIs

Custom models for additional languages are available on request.*

Saigen’s Text-to-Speech (TTS) technology converts text into high-quality, natural-sounding synthetic speech in real time. Our multilingual models currently support English and Afrikaans, with isiZulu, Setswana, and Sesotho in active development.

Key Features:

  • Custom voice models tailored to your brand’s tone, pace, and pronunciation

  • Natural pacing and pronunciation for improved user engagement

  • Multilingual support: English, Afrikaans, and more South African languages coming soon

  • Scalable B2B-ready APIs

Custom models for additional languages are available on request.*

Highly Accurate

Highly accurate speech recognition is one of the most exciting results of the Deep-Learning revolution. Until recently, such accuracy could only be achieved within severely restricted domains, by limiting the ‘vocabulary’ of words to be recognised.

Unrestricted domains

Through the magic of Deep Learning, we can now recognise virtually unrestricted domains with sufficient accuracy to support commercial applications.

Customer customisation

The large-vocabulary recognisers that we develop are optimised for our customers’ specific needs. See some examples below.

Specialising in under-resourced languages
Accurate audio and video transcriptions
Customised keyword recognition
Specialising in under-resourced languages
Accurate audio and video transcriptions
Customised keyword recognition

Our Pricing

TRANSCRIPTION

Get up to 60 minutes free. After that, credits can be purchased directly within the platform.

  • Punctuation & Capitilisation

  • Contact us for real-time transcription

  • Five languages supported, with more in the pipeline

REAL-TIME TRANSCRIPTION

$0.40 per audio hour

*Minimum volumes apply
  • Trained on conversational data. Fine-tuned with
    your data
  • Initial results in milliseconds
  • Enables real-time ASR for chatbots and
    conversational AI

MEDIA MONITORING

$2000 / month
OR
$1650
/ month

for a 12-month contract

  • Includes 17,500 hours of audio

  • Custom keyword recognition

  • Custom language models available at $2675

SPEECH ANALYTICS

For a custom quote, kindly reach out to Callbi
  • Multilingual code-switching

  • Easy to deploy, use, and afford

  • ISO27001 certified

TEXT-TO-SPEECH

For a custom quote, kindly get in touch
  • Natural-sounding speech converted from text

  • Multilingual technology

  • Real-time results

Our Pricing

TRANSCRIPTION

Get up to 60 minutes free. After that, credits can be purchased directly within the platform.

  • Punctuation & Capitilisation

  • Contact us for real-time transcription

  • Five languages supported, with more in the pipeline

MEDIA MONITORING

$2000 / month
OR
$1650
/ month

for a 12-month contract

  • Includes 17,500 hours of audio

  • Custom keyword recognition

  • Custom language models available at $2675

SPEECH ANALYTICS

For a custom quote, kindly reach out to Callbi
  • Multilingual code-switching

  • Easy to deploy, use, and afford

  • ISO27001 certified

TEXT-TO-SPEECH

For a custom quote, kindly get in touch
  • Natural-sounding speech converted from text

  • Multilingual technology

  • Real-time results