论文标题
为了自动化AI操作生命周期
Towards Automating the AI Operations Lifecycle
论文作者
论文摘要
当今的AI部署通常需要在模型生命周期的操作阶段进行大量的人类参与和技能,包括预发行测试,监测,问题诊断和模型改进。我们提出了一套可用于提高AI操作中自动化水平的能力技术,从而降低了所需的人类努力。由于人类参与的共同来源是需要评估部署模型的性能,因此我们专注于进行性能预测和KPI分析的技术,并展示如何在典型的AI操作管道的关键阶段中使用它们来改善自动化。
Today's AI deployments often require significant human involvement and skill in the operational stages of the model lifecycle, including pre-release testing, monitoring, problem diagnosis and model improvements. We present a set of enabling technologies that can be used to increase the level of automation in AI operations, thus lowering the human effort required. Since a common source of human involvement is the need to assess the performance of deployed models, we focus on technologies for performance prediction and KPI analysis and show how they can be used to improve automation in the key stages of a typical AI operations pipeline.