Amazon’s cloud computing arm, AWS, has introduced a new AI assistant named ‘Amazon Q’.

Amazon Q is designed to function as an AI assistant, enabling users to ask specific questions about their company’s data. This tool is particularly beneficial for employees who need quick access to information, such as company guidelines or understanding complex code, without the hassle of navigating through extensive documentation.

Currently, Amazon Q is integrated with Amazon Connect, AWS’s service for contact centers, and is set to expand to other AWS services. These include Amazon Supply Chain for tracking supply chain management and Amazon QuickSight, a platform for business intelligence. The AI assistant is trained on 17 years of AWS knowledge, suggesting it can provide highly relevant and accurate information for AWS users.

The AI assistant works by connecting to Amazon Bedrock, AWS’s repository of AI models, which includes advanced models like Meta’s Llama 2 and Anthropic’s Claude 2. Customers have the flexibility to choose the model that best fits their needs, connect to the Bedrock API, and deploy Amazon Q to adapt to their specific data, policies, and workflows.

A key feature of Amazon Q is its commitment to privacy and security. AWS CEO Selipsky emphasized that questions asked on Amazon Q would not be used to train any foundational models, ensuring that sensitive business data remains confidential. Additionally, the AI assistant recognizes security parameters set by customers, restricting access to sensitive information based on employee authorization levels.

Amazon Q’s deployment varies across AWS services. For instance, in Amazon Connect, Q operates in real-time, providing contact center agents with relevant answers during customer calls, thereby enhancing efficiency and customer service. Amazon Q’s current pricing model for Amazon Connect users starts at $40 per agent per month, with a no-charge trial period running until March 1, 2024.