Amazon Q imagines an AI future of work

From Accenture to BMW, major corporations are lining up to try a new kind of generative AI-powered assistant, aimed at employees, but with security and privacy at its heart.

Amazon Q, which helps ordinary individuals in an office environment to get answers to questions, solve problems, generate content, and take actions using company data and expertise, was announced on Tuesday during the opening of the Amazon Web Services (AWS) re:Invent 2023 conference in Las Vegas by CEO Adam Selipsky.

He said it could “easily chat, generate content and take actions. It’s all informed by an understanding of your systems, your data repositories and your operations.”

In a company statement, AWS said it will “streamline tasks, accelerate decision making and problem solving, and help spark creativity and innovation at work”.

Aimed squarely at large enterprises with big work forces, it is designed to stringent enterprise requirements, but at the same time can personalise  interactions with individual users based on an organisation’s identities, roles, and permissions. 

Unlike ChatGPT, which is known to use content posted on its platform to continue training itself – and sometimes includes that content in responses to other users – Amazon Q will not use business customers’ content to train its underlying models. This addresses a major hurdle to enterprise use of AI, following numerous horror stories of confidential information being leaked accidentally via chatbots.

AWS said in a statement: “Generative AI chat applications have captured the public’s imagination and helped people understand what is possible, but there are still barriers that prevent people from using these solutions at work. Specifically, these chat applications do not know an organisation’s business, data, customers, operations, or employees—the work they do, who they interact with, what information they use, and what they can access. 

“Additionally, these solutions were not initially built with the security and privacy features that organizations need for employees to safely use them in their day-to-day work. This has led to companies adding these features to their assistants after they were built, which does not work as well as incorporating security into the assistant’s fundamental design.”

Selipsky said in his keynote that AWS was harnessing generative AI in three layers of the AI stack, namely infrastructure, tools, and applications. 

Dr. Swami Sivasubramanian, AWS vice president of data and AI, said on Tuesday: “Generative AI has the potential to spur a technological shift that will reshape how people do everything from searching for information and exploring new ideas to writing and building applications.” 

Amazon Q promises to give developers access to powerful capabilities to solve common challenges, further simplifying application development and maintenance. The benefits include developing features faster and making updates to code in minutes instead of days.

Among other, Amazon Q has been built into the Amazon Connect cloud contact centre platform. Rather than replace contact centre agents, it can give them the ability to become high performers by having the information they need to respond to customers quickly and accurately. 

“Amazon Q in Connect detects customer issues based on the real-time conversation between the customer and agent, and automatically provides the agent responses, suggested actions, and links to relevant articles,” said the AWS statement. “By empowering agents to address customer needs across a broad range of topics without assistance from supervisors, Amazon Q in Connect increases customer satisfaction while reducing agent training, resolution time, and cost. 

“For example, Amazon Q could detect a customer is contacting a rental car company to change their reservation. Amazon Q would then generate a response the agent could send detailing the company’s change policies and guide the agent through the step-by-step process of updating the reservation.”

* To learn more about Amazon Q, visit aws.amazon.com/q.