宇芽智能专注于AI音箱解决方案,为企业客户提供定制化的智能语音交互产品。

Deployment

Multi-form privatization and hybrid deployment

Edge offline + regional cloud hybrid, lightweight image one-click deployment, meeting the network and compliance requirements of different countries and customers.

Privatization deployment

Edge offline

Model and resource localization, available offline; OTA incremental updates, saving bandwidth.

Regional Cloud

Deployed by country/region, with computation and data processed locally to meet data sovereignty requirements.

Hybrid Deployment

Edge-first, cloud as backup; flexible switching between edge and cloud based on computing power and cost.

On-premises deployment
On-Premises

Both systems and large models can be deployed locally.

  • Localization of voice/multimodal/business services, available offline.
  • LLM privatization (vLLM/llama.cpp/TensorRT-LLM, etc.)
  • Supports GPU scheduling and inference acceleration for AIGC workloads.
  • Cloud-isomorphic interface, allowing seamless switching between edge and cloud at any time.
vLLM TensorRT-LLM llama.cpp OpenAI compatible
Deployment mode
  • Container images (K8s/Compose) and offline installation packages
  • Two modes: hardware appliance or hardware platform licensing
  • Built-in monitoring and logging, observable out of the box
Maintenance support
  • Remote OTA, Gradual Rollout, and Rollback
  • SLA/Emergency Response, On-site Support Optional
  • Localized Language and Time Zone Support

Need Customized Delivery and Deployment Solutions?

Provides Integrated Global Delivery Support from Hardware to Software