AI and big-data solutions for power-grid operations and energy dispatch
广州佰聆 builds AI and data-infrastructure products for China's power-grid sector. The stack spans PyTorch, TensorFlow, MindSpore, and PaddlePaddle — indicating deep investment in model training across multiple frameworks — alongside production systems (Kafka, RabbitMQ, Elasticsearch, Redis). Active projects focus on grid-operation forecasting, LLM research for energy-domain scenarios, and time-series prediction, suggesting the company is transitioning from traditional big-data platforms toward applied AI for grid management and dispatch optimization.
广州佰聆 is a 32-person software company headquartered in Guangzhou serving China's power industry. The business spans three dimensions: big-data infrastructure for grid operations, large-language-model research and deployment for energy-sector use cases, and field solutions for grid forecasting and dispatch planning. Engineering and data roles dominate the org (39 of 32 total headcount), with consulting and operations providing domain support. Current challenges center on LLM deployment complexity, architecture planning for integration, and cost control in energy-sector implementations.
PyTorch, TensorFlow, MindSpore, and PaddlePaddle. The company also uses Transformers library for model development and maintains active research into vertical-domain LLMs for power-grid scenarios.
Power-grid forecasting, LLM research for energy scenarios, grid-operation time-series prediction, and AI-powered dispatch solutions. Current focus includes model optimization, data preprocessing for power-industry computing, and scenario application development.
Other companies in the same industry, closest in size