KETs 2026 EU Brokerage Event

4–5 Feb 2026 | Karlsruhe, Germany

Register
Register
Register

Project cooperationUpdated on 21 January 2026

Real-Time GenAI Audio Brick for Edge & Telecom Infrastructures

CEO / CTO at CandyVoice

Strasbourg, France

About

CandyVoice proposes a Real-Time GenAI Audio Brick designed for edge computing and telecom infrastructures, addressing the growing need for robust, low-latency and trustworthy voice processing in mission-critical environments.

The solution combines advanced DSP and AI models to deliver real-time audio enhancement and voice intelligence under strong operational constraints (latency, bandwidth, power, security). It is designed to be deployed close to the signal source (edge devices, gateways, embedded systems) or within telecom network elements, reducing dependency on cloud processing.

The audio brick integrates modular capabilities such as:

  • real-time noise reduction and speech enhancement in degraded conditions,

  • packet loss concealment and signal restoration for unreliable networks,

  • privacy-preserving voice processing (anonymisation, transformation),

  • optional voice integrity and deepfake detection layers.

The technology is already validated on TRL 6–7 demonstrators, with SDKs suitable for C/C++ integration, embedded platforms and edge architectures. It supports privacy-by-design and AI Act–ready approaches, making it suitable for regulated and safety-critical domains.

CandyVoice seeks project cooperation within Horizon Europe (RIA or IA) to:

  • integrate this GenAI audio brick into larger edge, telecom or critical communication systems,

  • jointly develop and validate use cases such as emergency communications, network operations, field maintenance, secure voice services or human–network interaction,

  • contribute a specialised audio/voice layer to multidisciplinary consortia.

Target partners include telecom actors, edge/embedded system providers, system integrators, research organisations and project coordinators addressing resilient, trustworthy and sovereign AI infrastructures.

------------
GenAI audio
Real-time voice processing
Edge AI
Telecom infrastructures
Low-latency AI
DSP + AI
Mission-critical communications
Privacy-by-design
AI Act
Horizon Europe
Embedded systems
Resilient networks

Stage

  • Planning

Topic

  • 3C Networks
  • AI-GenAI /Data/Robotics
  • Virtual Worlds

Type

  • Partner seeks Consortium

Organisation

CandyVoice

SME

Strasbourg, France

Similar opportunities