ameau01/synthesized-cloud-optimization-recommendations

该数据集名为“合成云优化建议”，包含18个场景，每个场景将云遥测数据与手工制作的优化建议配对。数据集用于训练模型或评估AI代理。每个场景包括多层遥测数据、描述部署基础设施的Terraform文件以及一个黄金标准建议。数据集围绕简单的输入-输出映射构建：输入是遥测数据和基础设施，输出是优化建议，说明需要更改的内容及其影响。数据集是合成的，遥测数据根据场景叙述程序化生成，建议是手工制作并验证的。数据集全程使用AWS词汇，如实例类型、服务名称和字段名称，使场景具体化而非供应商中立。文件夹结构包括场景摘要文件和每个场景的详细文件，覆盖单层、跨层、无操作、诊断延迟等不同优化情况。数据集可用于训练或微调模型，将云遥测映射到优化建议，或评估AI代理在云优化推理上的表现。The dataset is named Synthesized Cloud-Optimization Recommendations and includes 18 scenarios that pair cloud telemetry with hand-crafted optimization recommendations. It is used to train models or evaluate AI agents. Each scenario has multi-tier telemetry, a Terraform file describing the deployed infrastructure, and a gold-standard recommendation. The dataset is built around a simple input-output mapping: the input is telemetry plus infrastructure, and the output is an optimization recommendation that specifies what to change and its impact. The dataset is synthesized, with telemetry generated procedurally to match each scenarios narrative, and gold recommendations hand-crafted and verified. It uses AWS vocabulary throughout, including instance types, service names, and field names, making the scenarios concrete instead of vendor-neutral. The folder layout includes a summary file and detailed files per scenario, covering single-tier, cross-tier, no-action, diagnostic deferral, and other optimization situations. The dataset can be used for training or fine-tuning models that map cloud telemetry to optimization recommendations, or for evaluating AI agents on cloud-optimization reasoning.