Amazon has announced the launch of Q, an AI-powered chatbot designed for AWS customers.
The announcement came during a keynote at Amazon’s re:Invent conference, revealing that Q has the capability to respond to inquiries such as “how do I build a web application with AWS.” Utilizing 17 years’ worth of AWS knowledge, it provides a list of potential solutions and reasons for considering each proposal.
AWS CEO Adam Selipsky stated, “You can easily chat, generate content and take actions. It’s all informed by an understanding of your systems, your data repositories and your operations.”
Organizations can configure Q by connecting it to and customizing it with their specific apps and software, like Salesforce and Amazon S3 storage instances. Q learns and indexes data and content about a business, including core concepts, product names, and organizational structure.
Users can ask Q to analyze various aspects, such as product features customers are struggling with, or upload a file and ask questions about it. Q draws on its data to provide relevant responses along with citations.
Q is not limited to answering questions; it can also take actions on behalf of users through configurable plugins, such as creating service tickets or updating dashboards. Users have the opportunity to inspect and verify any actions before they are executed.
Integrated in the AWS Management Console as well as existing chat and business apps like Slack, Q can understand app workloads and suggest AWS solutions and products based on specific requirements.
Selipsky provided an example where Q would assist in selecting the best EC2 instance for a particular app, considering performance and cost considerations.
Additionally, Q can troubleshoot network connectivity issues and ties in with CodeWhisperer, Amazon’s service for code generation and interpretation.
Q is being integrated into first-party products such as QuickSight and Amazon Connect, enhancing capabilities in business analytics and customer service.
Amazon leveraged Q internally to upgrade around 1,000 apps from Java 8 to Java 17 and to test those apps within just two days.