Techmo TTS Documentation

Available voices

Polish - Masza

Play

Play

Polish - Michal

Play

Play

German - Thorsten

Play

Play

English - Cori

Play

Play

Spanish - Tomas

Play

Play

... more examples

Main features

• Minimal latency streaming,
  starting from 70 ms

• All-in-one platform for easy creation and customization of synthesized prompts,
  with granular control over text and audio parameters at hand

• Perfect adjustments for proper audio response formatting with SSML tags

Even more features…

• Customizable resources, allowing personalized
  pronunciations of specific words and proper nouns

• International Phonetic Alphabet support, enabling precise
  control over pronunciation details

• Extensive library of expressive non-verbal sounds,
  enriching synthesized speech with laughs, sighs and more

• Available Languages:
  German, Spanish, English, Polish

• Wide range of sampling rates (8-22 kHz) for PCM16, OGG VORBIS, OGG OPUS, A LAW, MU LAW formats

Technology

• Ready to use C++ client implementation

• Ready to use Python client implementation

• Utilizing gRPC technology

• Specifically tailored for the MRCPv2 protocol

• TLS/mTLS encryption for confidentiality and integrity of data

On-Premises

• Container management with Kubernetes technology

• Containerization management with Docker technology

• No needs for GPU cluster to generate company voice

• > 2 simultaneous streaming channels from one cpu core

• Over 1000000 hours per year in customer service

• Billing by channels or characters and minutes

• Monitor and system performance analysis with Prometheus

• Working on Red Hat family OS distributions

• Simple to be used on Ubuntu

First steps

• Examples
  • Trying Different Variants for Different Interpretation
  • Boosting Positivity with Prosody
  • Humanizing Synthesis with Sound Icons, Breaks and Tweaking Punctuation
  • Increasing Intelligibility with Prosody
  • Making Speech More Natural with Randomized Breaks and Automated Normalization
  • Using Phonetic Alphabet
• Architecture
• How to use TTS
• Synthesis guide

Client applications

• Demo website
  • Voice Recognition and Synthesis Testing Platform
• TTS Studio
• C++ CLI client
  • Docker usage
  • Local instance usage
• Python CLI client
  • Local instance usage
  • Manual usage
  • Docker usage
  • Usage

TTS gRPC API

• gRPC API
  • Functions Definitions
  • Requests and Responses Definitions
  • Enumerations

On-premises configuration

• TTS system overview
  • System requirements
  • System architecture
  • Directory structure of the deployment package
• Startup options and configuration
  • License configuration
  • Resources configuration
  • Runtime limits configuration
  • Lexicons configuration
  • Voices configuration
  • Recordings configuration
• Docker setup
  • Loading docker image
  • Service options and configuration
  • Licensing data generation
  • TLS authentication
  • Starting the Service
• MRCP environment
  • MRCP proxy
  • Configuration of MRCP environment

Management and tracking services

• Service logs
  • gRPC health-check
• REST status interface
• Event monitoring with Prometheus

References

• Partners