Orca Security 基于 Amazon Bedrock 构建自然语言安全查询
原标题：Natural language security queries at Orca Security using Amazon Bedrock
AI 摘要
Orca Security 与 AWS 合作，基于 Amazon Bedrock 构建了自然语言安全查询代理 Discovery AI，可将用户的自然语言问题转换为 Sonar 查询。该系统采用九阶段混合流水线，结合 RAG 与 schema 感知查询构建，并针对不同阶段选用不同基础模型。相比最初将完整 schema 塞入单个提示的基线方案，Orca 将查询成本降低 90%、输入 token 减少 93%、延迟降低 75%，实现每天处理数千次查询、响应时间约 15 秒。
正文节选
Natural language security queries at Orca Security using Amazon Bedrock Security analysts need fast answers about their cloud infrastructure without deep familiarity with domain-specific query languages. Orca Security’s tool covers over 900 cloud models (the resource types Sonar tracks, such as an Amazon Elastic Compute Cloud (Amazon EC2) instance or an Amazon Simple Storage Service (Amazon S3) bucket). Querying that data required precise model names, nested relationship navigation, and selectio