最新资讯
stackoverflow.blog 欢迎来到芯片时代:半导体制造、AI 数据中心与芯片设计的未来 www.youtube.com AI Agent 总犯同样的错:CentralAgent AI 将记忆重构为经验系统 pub.towardsai.net Netflix 科学家建议重采样 LLM 评判:5 次仅将误差棒缩小 13% pub.towardsai.net Jev 作为 LLM 前的静默层:构建提示感知的模型路由器 reddit.com OpenAI 称 Astra 仅满足 10% 安全任务标准 dev.to FreeToken:在自有 GPU 上运行前沿 MoE 模型及其内存需求 dev.to 当 AI Agent 说“完成”时,究竟完成了什么? dev.to 部署 Underdog Bark 3.0 Pro:逐字转写、自定义后端与分块设置 dev.to 为艰难对话构建 AI:五个 Agent、一条安全正则与一位盲评监督者 techcrunch.com Palo Alto Networks 洽谈以 5 亿美元收购 AI Agent 安全初创公司 Clutch theinformation.com 黄仁勋被要求就涉华间谍指控向国会作简报 www.infoq.com Edera 推出 Object Capability Bounded Authority,试图将凭证与 AI Agent 工作负载隔离
多元动态
登录 注册
智能体 aws.amazon.com · @AWS Machine Learning Blog

Amazon Quick 桌面端正式可用

原文 我的收藏

AWS 宣布 Amazon Quick 桌面应用在 macOS 和 Windows 上正式可用,同时为 iOS/Android 移动端新增活动聚合视图,将邮件、日历、CRM 和消息整合为统一优先级列表,由 agent 在后台处理常规事项。Quick 运行于 AWS,数据保留在客户环境内,对话保持私密,可通过 Amazon CloudWatch 和 AWS CloudTrail 获取完整审计记录,并内置 HIPAA、FedRAMP、SOC 2 和 ISO 27001 合规认证。官方称制造、医疗和体育行业客户在预览期已使用该产品。

AI推荐理由 AWS 官方博客宣布 Amazon Quick 桌面端(macOS/Windows)正式可用,并新增 iOS/Android 活动聚合视图;正文给出数据驻留、CloudWatch/CloudTrail 审计及 HIPAA、FedRAMP、SOC 2、ISO 27001 合规等具体能力,属可核验的官方产品发布。但内容以营销叙事和客户场景为主,缺少定价、性能指标与独立评测,且‘客户已在预览期使用’等表述无量化证据,故兴趣分中等。

62

正文

Your teams get an AI assistant that handles real work while your data stays in your environment and your conversations stay private Today, the Amazon Quick desktop application is generally available on macOS and Windows. We’re also adding a new activity feed to the mobile experience on iOS and Android that consolidates email, calendar, CRM, and messaging into one prioritized view, showing your teams the decisions that need them while agents resolve the routine items in the background, so they spend less time sorting through updates and more time on high-value work. Enterprise IT organizations face a compounding problem: the tools your teams need to stay productive are multiplying, and so is the shadow AI risk that comes with them. When people work around approved systems to get things done faster, governance breaks down, data leaves controlled environments, and IT loses visibility. The answer is not to restrict AI access but to give teams an enterprise-grade AI assistant built for work, on infrastructure IT already controls. Quick runs on AWS, the same infrastructure behind the most security-sensitive workloads on the planet. Your data stays in your environment, your conversations remain private, and everything is auditable. Full audit trails are available through Amazon CloudWatch and AWS CloudTrail. The compliance certifications your security team requires (HIPAA, FedRAMP, SOC 2, and ISO 27001) are built in from day one. Customers across manufacturing, healthcare, and sports have been using Quick on desktop during preview—asking complex questions in natural language and getting grounded, trustworthy answers in seconds, cutting time-to-insight significantly while keeping governance and accuracy built in from the start. For organizations managing knowledge workers across hundreds of roles and workflows, the capacity equation matters too. Business goals keep growing, but the hours in a day do not. Most of the time that should go to high-judgment work gets consumed by pulling information together, chasing follow-ups, and assembling deliverables. Quick changes that equation by providing a thought partner that knows your context, builds what you need, and takes action on your behalf. Quick also gives your team a shared workspace where the dashboards, agents, and automations one person builds are available to the whole team. Here is what Quick looks like across a real workday, from delegating meeting prep on a laptop to triaging priorities on a phone and following up after the workday ends. Here, an enterprise account executive at a manufacturing company uses Quick to dramatically simplify prep for a key customer meeting—typing a single request and letting Quick assemble the brief in the background Focus on what matters, not what’s loudest Most work tools generate a firehose of notifications. Quick’s activity feed does the opposite: it consolidates signals from email, messaging, CRM, and calendar into one prioritized view. Items agents resolve on their own disappear. What remains is a short queue of things only you can act on, and the feed gets smarter over time. Quick learns your relationships, priorities, and patterns, so it surfaces what matters most, not just what arrived most recently. Here, the same rep triages her morning from her phone, approving an addition to her meeting brief and responding to a pipeline review request for her VP, without opening another app. For teams operating at enterprise scale, that prioritization is what turns a firehose of updates into a clear path forward: “At Southwest Airlines, we’re building agentic AI tools and autonomous agents to streamline operations across our 70,000+ Employees and enhance the experience for the millions of Customers we serve every day. Amazon Quick Desktop gives our teams the ability to ask complex questions in natural language and get grounded, trustworthy answers in seconds instead of waiting on ad hoc report requests. Our developers build once and deploy intelligent experiences across the enterprise, with governance and accuracy baked in from the start. Quick cuts time-to-insight significantly and is already powering high-impact use cases across the organization including market analytics. It’s become the foundation of how we put AI-driven intelligence directly into the hands of our People.” -Justin Bundick, VP Technology Intelligence Platforms, Southwest Airlines From meeting to momentum Most AI tools answer questions. Quick completes tasks by synthesizing information across systems, drafting deliverables, updating records, and executing follow-ups on behalf of your team. Teams do not get a summary of what they could do. They get finished output, ready for review. Here, she picks up on her laptop where she left off on mobile—reviewing the completed work in mission control and sharing the finished brief with her account team. Early customers point to that shift from answers to finished work as the thing that changes how their teams operate: “We’ve been evaluating Amazon Quick ahead of its general availability launch, and the adoption was easy. Clean design, intuitive experience, and the knowledge graph and memory adapt to the way you work. You can create agents that run asynchronously on your behalf and go from concept to working prototype in a fraction of the time. Quick is becoming a force multiplier that frees you to focus on higher-value work.” — Chuck Metturdharma, VP & Chief AI Officer, LabCorp The work follows through, even when you don’t have time Quick is built on the tools, systems, and infrastructure your organization already runs on. There is no migration, no new ecosystem to adopt. It connects to what you have and starts delivering value on day one. Here, after the meeting, she wraps the day from her phone, having Quick refine and send a customer follow-up, then queue up the next morning’s materials for engineering. Every decision stays yours while Quick handles the rest. Whether you’re an IT leader, account executive, a program manager juggling cross-functional launches, or a finance lead closing the books, Quick works the way you do, across every device, on your schedule. You bring the judgment; Quick gives your teams the capacity to match their goals. That capacity shows up differently for every team, but the pattern is the same: people spend less time on setup and more time acting on what they know: “Amazon Quick allows our domain experts—the people who know golf and our data best—to do work that used to sit in a technical backlog. We’ve prototyped working systems in days instead of weeks. When an idea proves out, they can hand our engineering team well-documented, validated requirements instead of a rough concept, so development starts from a proven foundation. And with it on mobile, they can act on an idea the moment it hits instead of waiting until they are back at a computer.” — Randall Kato, VP Golf Technology, PGA TOUR These are the kinds of results customers are already achieving with Quick, and they represent what becomes possible when your team can delegate the routine and focus on the work that moves the business. Download Quick today Spencer Martenson Spencer has spent six years at AWS translating complex technology into stories that resonate with customers. He previously led product marketing for SAP on AWS, helping enterprises modernize their critical ERP workloads in the cloud. Now he’s focused on Amazon Quick — building GTM strategies that help customers work smarter by taking advantage of the latest advancements in AI. Spencer lives in Seattle, spending his spare time pursuing rare birds and training his dog. Chris Lott Chris is a Principal GenAI/ML Specialist Solutions Architect on the Amazon Quick team. He has over 25 years of enterprise software development experience. Chris enjoys gardening, cooking, aerospace/general aviation, and traveling the world. Ramon Lopez Ramon is a Principal Solutions Architect for Amazon Quick. With many years of experience building BI solutions and a background in accounting, he loves working with customers, creating solutions, and making world-class services. When not working, he prefers to be outdoors in the ocean or up on a mountain.

来源报道 · 3 篇 · 以下链接均前往原始媒体

AWS Machine Learning Blog Amazon Quick 桌面端正式可用 09-11 · 原文发布
Spencer Martenson Amazon Quick 桌面端正式可用 09-11 · 原文发布
Spencer Martenson Amazon Quick 桌面端正式可用 09-11 · 原文发布
← 返回精选动态 原文

更多内容