Flexibility, scalability, and adaptability
Our proprietary voice module KIDOU comprises a wide range of AI components for processing speech and text. It overcomes challenges such as accurate speech recognition in regional dialects and noisy environments, processes specialized technical terms, and can handle extensive columns of numbers.
KIDOU voice toolkit with powerful AI components
Our AI components are designed to be flexible, scalable, and adaptable in order to meet the requirements of a wide variety of business use cases.
Noise reduction
High levels of noise or background noise are expected in the operating environment.
High-precision noise reduction tailored specifically to your environment to ensure optimal voice quality.
Technologies: Signal Processing, MEL, Deep Learning
Voice recognition / Voice Activity Detection
Precise detection of speech activity, regardless of background noise.
Use case: Automatic transcription of meetings where only spoken content is to be recorded.
Technologies: Machine learning, signal processing.
Text-to-speech / tts
The system should also respond in natural language, e.g., in situations where it is not possible to look at a screen, such as in a car, in an operating room, etc.
Generates natural speech output that can be customized to your brand identity.
Use case: Personalized customer communication, e.g., automatic response to customer inquiries by phone.
Technologies: Deep learning, natural language processing
Dialogue guidance
Supports natural and effective conversations between users and systems, tailored to your specific use cases.
Use case: Customer service chatbot that can handle complex inquiries.
Technologies: Natural language understanding, dialogue management, deep learning
Speaker identification
Reliably identifies individual speakers and enables personalized interactions.
Use case: Assigning different speakers in transcripts, e.g., of court hearings or meetings, authenticating users in voice control systems.
Technologies: Machine learning.
Wakeword / Hey KIDOU
Activates voice systems precisely and reliably, prevents transcription when deactivated, and significantly extends battery life.
Use case: Activation of a voice assistant using a specific word.
Technologies: Deep learning, signal processing.
Speech-to-text / stt
Converts spoken words into text formats and offers precise speech recognition. This is included in almost every use case. Through special training, our component recognizes the specific technical terms, dialects, accents, and phrases of your domain and is extremely robust against disruptive ambient noise.
Use cases: Free dictation in any application, documentation of defects during inspections and maintenance, recording of diagnoses or treatments, transcription of court hearings or meetings.
Technologies: Deep learning.
Matcher / Speech-to-structure
Quick and easy form filling and command control
Recognizes and extracts structured information from text, even if it is incorrect. This information is then structured and made available in a uniform format, e.g., JSON, for further processing.
Use cases: Inspections and/or maintenance to identify errors with location, severity, and components involved.
Technologies: Deep learning.
Text and Document classification
Analyzes large amounts of text data to classify it and extract relevant information.
Use case: Automated categorization of customer feedback.
Technologies: Machine learning, natural language processing.
Question & Answering
Enables precise answers to complex questions from extensive documents or databases. If required for intellectual property (IP) or data protection reasons, also with your own large language model (LLM) that can be operated on-premises at your company.
Use cases: Internal system that answers questions about procedural documentation, automated customer support, e.g., answering FAQs.
Technologies: Large Language Models, Deep Learning, Natural Language Processing
Voice command for apps
With the components described, your existing app can be made voice-enabled, allowing your customers and employees to work with their familiar app and also use voice for control and input.
Use case: Company-owned app for documenting errors during inspections and/or maintenance
Sentimental analysis
Recognizes and understands the mood and emotions in written or spoken text.
Use case: Responding in dialogue depending on customer mood in a telephone voicebot.
Technologies: Machine learning, natural language processing.
Do you have further questions about our AI components or would you like a free consultation?
Then we look forward to receiving your contact request.
Self-developed & independent
Seamless KIDOU integration and maximum flexibility
All KIDOU AI components were developed in-house by KENBUN and are completely independent of products or services from other manufacturers. This allows us to offer you seamless integration and maximum flexibility to meet your individual requirements.
Welcome to KENBUN – your reliable partner for customized voice assistance systems with KIDOU.
