Anders Swanson 是 Oracle Database 的开发者布道师，在 AI Engineer 演讲中讨论从上下文到记忆：智能体需要真正的记忆层。人物身份依据原发布方署名，职务按节目发布时记录。
Anders Swanson is a speaker from Oracle Database in the indexed AI Engineer talk. The episode discusses From Context to Memory: Your Agents Need a Real Memory Layer. Affiliation is recorded from the publisher’s speaker credit at publication.
介绍来源
原发布方人物署名与介绍 · 核对于 2026-10-10。依据原节目发布时的介绍。
图片来源与许可 · 参考图使用许可未确认
AI 辅助插画 · 参考照片：AI Engineer · 参考图使用许可未确认 · 依据已核对姓名的来源页面图像，由 AI 绘制 B 细线彩铅头像；调整背景与肩部构图。展示图缩放至最长边 640 像素并转为 WebP。 · 参考图：Anders Swanson — 原发布方人物照片
公开发言
3原话逐字核对过；转述不加引号redaction is very important as the first step because before I do embeddings, before I do enrichment, I want to strip out PII. I don't want secrets or irrelevant weird information slipping into the memory layer. That's a recipe for disaster. So, we do redaction first.
And this this form of retra retrieval can fuse different search methods. So, I can take vector similarity search, I can take a full-text keyword search, I can take relational entities, I can do graph traversal, I can take other feedback signals like importance uh uh and usefulness of the memory to yield more comprehensive results than one specific query branch alone.
Anders Swanson 指出，智能体运行时可通过 Hook、MCP 或技能机制将记忆层作为外部功能调用，从而将状态存储与模型推理明确解耦。