Amazon Web Services (AWS) has debuted a chat-based AI assistant designed to help the enterprise worker by integrating the mishmash of tools they deploy in their daily labors.
AWS is deploying Quick using a “Trojan Horse” strategy. By offering an inexpensive seat license to employees (mimicking Slack’s grassroots success), AWS aims to encourage wider corporate adoption of its AI stack–at a considerable increase in cost.
But this approach also solves a major challenge for enterprise AI: corporate access. Quick functions best when users grant it access to their files and applications, a scarce resource for the data-hungry AI.
Unlike similar cloud services from Microsoft Copilot or the Google Gemini Enterprise Agent Platform, Quick is not partial to its own productivity stack. With connectors, Quick ties together corporate warriors’ hodge-podge of apps, including Office 365, Slack, Jira, Gmail and Outlook and about 50 others.
Shadow AI for the Corporate Warrior
Following Slack’s formula of success through grassroots adoption, this “AI personal assistant” is being marketed directly to the employees themselves.
“Most of us still spend more time hunting for information at work than using it to get our jobs done effectively,” wrote Jigar Thakkar, Amazon vice president of agentic AI for business, in a blog post introducing the technology.
“You need AI that actually understands you—how you work, who you work with, your data, workflows, and information across every system.”
New users are encouraged to sign up with their personal email credentials (although they can sign up with their AWS credentials as well).
Quick Consolidates Business Apps
The desktop assistant provides the worker with a single workspace where they can issue agentic directions to their communications apps. If you are working on a presentation with a colleague, you can email the colleague, schedule a meeting and do research – all from the command line, or voice command.
“Most AI tools only work within their own vendor-specific ecosystem and can only help with a fraction of your work, Quick is built to break you free from those walled gardens,” Thakkar wrote.
Quick is model-agnostic, though by default it uses AWS’ own Nova 2 Lite model – with a generous 1 million token context window for hardy reasoning. MCP serves as an AI universal remote, allowing the user to dial in other AI tools such as Claude Cowork or Kiro CLI.
Although you can run a web version, the desktop app indexes the files on the user’s machine, giving the AI more domain-specific information about the user’s job and daily tasks, which can, with additional AI resources, be used to build a knowledge graph of their concerns.
AI’s Way into the Enterprise
For Quick to gather all this knowledge, the user must supply the access to their files and apps, thus solving one of the thorniest challenges for AI: corporate access.
Running agents enterprise-wide necessitates a quagmire of granting and managing permissions across myriad apps and files. Initially serving employees first is the easiest way for an AI to access files, execute merge requests, copy email, and do all those things that it needs to do to provide value.
“The scarce asset in enterprise AI may be shifting from intelligence to permission,” noted AI observer Jaya Gupta, in an X article.
Thus far, 3M, GoDaddy, AstraZeneca, BMW, the NFL, Southwest Airlines, as well as Amazon itself, are all testing Quick, according to the company.
In a press release, a New York Life executive enthused how a single conversational agent can provide answers without the need for the user to pull reports or contact analysts.
Pricing includes a free testing edition and an individual plan at $20/month (billed annually). Enterprise professional accounts cost $20/month per seat or $40/month per seat, and both require an additional $250 monthly “infrastructure fee” for advanced features like corporate knowledge graphs and organization-wide data governance.
