Amazon is launching an AI-powered chatbot for AWS customers called Q.
Revealed during a keynote at Amazon’s re:Invent conference in Las Vegas this morning, Q – starting at $20 per user per year – can answer questions like “how do I build a web application using AWS?” Trained in AWS knowledge for 17 years, Q will offer a list of potential solutions along with reasons you can consider her suggestions.
“You can easily chat, create content and take action [with Q]AWS CEO Adam Selipsky said on stage. “All of this is informed by an understanding of your systems, data stores and operations.”
AWS customers configure Q by connecting it to — and customizing it with — enterprise-specific apps and software like Salesforce, Gmail, and Amazon S3. Q indexes all linked data and content, “learning” aspects about a business, including its organizational structures, key concepts, and product names.
From a web app, a company can ask Q to analyze, for example, what product features its customers are experiencing and possible ways to improve them — or, a la ChatGPT, upload a file (a Word document, PDF, computer leaf and like) and ask questions about this file. Q relies on its connections, integrations and data, including business-specific data, to find answers along with reports.
Q goes beyond simply answering questions. The assistant can take actions on behalf of a user through a set of configurable plugins, such as automatically creating service tickets, notifying specific teams in Slack, and updating dashboards in ServiceNow. To avoid mistakes, the chatbot has users inspect the actions they are about to take before they are executed and link the results for validation.
Accessible from the AWS Management Console and the aforementioned web app, as well as from existing chat apps like Slack, Q has a thorough understanding of AWS and the various products and services available through it, as you might imagine. Amazon says it can understand the nuances of application workloads on AWS, recommending AWS solutions and products for applications that run for only a few seconds, for example, or access storage very infrequently.
In the tent, Selipsky gave the example of an application that needs high performance for video encoding and transcoding. Asked about the best EC2 case for the application in question, Q would present a list considering performance and cost, Selipsky said.
Q can also troubleshoot issues such as network connectivity issues, analyze network configurations to provide remedial steps.
And Q connects to CodeWhisperer, Amazon’s service that can generate and interpret code. Within a supported IDE (eg, Amazon’s CodeCatalyst), Q can generate tests for software benchmarking — based on knowledge of a customer’s code. Amazon Q can also create a draft plan for implementing new features in software or code transformation and upgrading code packages, repositories, and frameworks—plans that can be refined and even executed using natural language.
Selipsky says Amazon used Q internally to upgrade about 1,000 applications from Java 8 to Java 17—and test those applications—in just two days.
Amazon is also integrating Q into first-party products, such as QuickSight, the company’s business analytics service. Q can provide visualization options for business reports, automatically reformatting them. Or it can answer questions about data in a report.
Q is also making its way into Amazon’s contact center software, Amazon Connect. Now, powered by Q, customer service agents receive suggested answers to customer questions with suggested actions and links to relevant support articles — without having to type those customer questions on a line of text. Q also generates a post-call summary that supervisors can use to track follow-up steps.
Amazon has emphasized several times that the answers Q gives – and the actions it takes – are fully controllable. In practice, this means that if users only return information that they are authorized to see. Admins can limit sensitive topics by filtering out inappropriate questions and answers where appropriate. And, at least by default, Q models—a combination of models from Bedrock, Amazon’s AI developer platform, including Amazon’s in-house family of Titan models—are not trained on a customer’s data.
In many ways, Q is like Amazon’s answer to Microsoft’s Copilot for Azure, which in turn was Microsoft’s answer to Duet AI on Google Cloud. Both Copilot for Azure and Duet AI on Google Cloud take the form of a conversational assistant for cloud customers, suggesting configurations for applications and environments and helping troubleshoot by identifying potential problems — and solutions.
But Q appears to be a bit more comprehensive — touching on a wide range of business intelligence, scheduling and configuration use cases. Ray Wang, founder and principal analyst at Constellation Research, believes it’s re:Invent’s “most important” announcement yet.
“It’s about arming developers with AI so they can be successful,” he said in a statement.
We’ll just have to see if it works as well as Amazon says.
