当前位置:首页 > 报告详情

利用 Oracle AI 数据平台加速 AI 代理开发 [LRN3463](1).pdf

上传人: Fl****zo 编号:970984 2025-11-08 31页 2.17MB

1、 Accelerating AI Agent Development with Oracles Data Platform Jean-Rene GauthierProduct Manager,Oracle AI Data PlatformOctober 15,2025Colin SchmidtVice President Engineering,Oracle AI Data PlatformThe following is intended to outline our general product direction.It is intended for information purpo

2、ses only,and may not be incorporated into any contract.It is not a commitment to deliver any material,code,or functionality,and should not be relied upon in making purchasing decisions.The development,release,timing,and pricing of any features or functionality described for Oracles products may chan

3、ge and remains at the sole discretion of Oracle Corporation.Safe harbor statement2Copyright 2025,Oracle and/or its affiliates|Confidential:Internal/Restricted/Highly RestrictedSpeakersJean-Rene GauthierProduct Manager,OracleColin SchmidtVice President Engineering,Oracle3Copyright 2025,Oracle and/or

4、its affiliates|Confidential:Internal/Restricted/Highly RestrictedGenAI adoption is climbing and fast!Copyright 2025,Oracle and/or its affiliates|Confidential:Internal/Restricted/Highly Restricted471%Use GenAI in at least one business functionSource:McKinsey&Company,2025 yet1%of company executives de

5、scribe their gen AI rollouts as mature5Copyright 2025,Oracle and/or its affiliates|Confidential:Internal/Restricted/Highly RestrictedSource:McKinsey&Company,2025 But ChatGPT is great!Copyright 2025,Oracle and/or its affiliates|Confidential:Internal/Restricted/Highly Restricted6Source:MIT NANDA,2025

6、For general purpose,adoption of GenAI through unofficial channels is very high and successful!Generic tools lack persistent memory and feedback loops Agent needs deep customization aligned to internal data Agent benchmarks need to be aligned with operational outcomes7Copyright 2025,Oracle and/or its

word格式文档无特别注明外均可编辑修改,预览文件经过压缩,下载原文更清晰!
三个皮匠报告文库所有资源均是客户上传分享,仅供网友学习交流,未经上传用户书面授权,请勿作商用。
根据《Accelerating AI Agent Development with Oracle's Data Platform》的内容,以下是全文关键点的概括: 1. **GenAI adoption trends**: 71% 的企业至少在一个业务功能中使用 GenAI,但只有 1% 的公司认为其 GenAI 部署成熟。 2. **AI Agent Development Challenges**: 通用工具缺乏持久记忆和反馈循环,需要针对特定流程进行深度定制。 3. **Oracle AI Data Platform Introduction**: Oracle AI Data Platform 通过结合企业数据与顶级 AI 模型和开发工具,帮助企业构建 AI 代理和应用程序。 4. **Platform Components**: 包括 AI 代理、数据与 AI 基础设施、AI 驱动的开发工具、Agentic 用户体验和开放湖屋架构。 5. **AI Agent Development Features**: 低代码和代码优先的 AI 开发、OCI 生成式 AI 与 Oracle 23ai 集成、生命周期管理、Fusion AI 代理工作室集成。 6. **Agent Flows**: 通过节点和关系定义的端到端代理应用程序,支持单代理、多代理系统和部分确定性执行。 7. **Guardrails and Monitoring**: 提供强大的默认保护措施,包括内容审查和统一 RBAC,以及全面的监控功能。 8. **Knowledge Bases**: 存储文档的向量表示,供代理搜索和检索相关文档。 9. **Development Roadmap**: 计划在 2025 年第三季度推出,并持续扩展功能,包括多代理支持、MCP 服务器、Fusion AI 代理工作室集成等。
Oracle平台揭秘" "企业AI应用,Oracle如何助力?" 构建智能未来"
客服
商务合作
小程序
服务号
折叠