MUXU
Edge AI & IoT Applications
On-device AI inference and IoT application services for embedded Linux.
Services
From model optimization to full IoT architecture — deploying AI and connected applications on embedded devices.
On-device LLM inference and deployment
Local LLMs for privacy-preserving, low-latency NLP without cloud dependency.
Model optimization and quantization
INT8/INT4 quantization, pruning, distillation for constrained devices.
Neural network accelerator integration
NPU, GPU, TPU integration for inference throughput and energy efficiency.
Sensor fusion and data preprocessing
Combining multi-sensor data into clean inputs for AI models.
Edge-to-cloud AI pipelines
Hybrid architectures: edge inference + cloud model updates and analytics.
IoT application architecture
End-to-end IoT design: protocols, data flow, device management.
Frameworks and runtimes
Open source frameworks and runtimes for deploying machine learning models and large language models on embedded devices.
Inference engines
Local inference engines for running large language models on edge hardware.
