Amazon QuickSight has AI agents that enable every user to perform the tasks of a data analyst and dive deep into available data. The agentic AI helps automate complex and multi-step analysis processes.
The BI platform QuickSight from AWS offers (business) end users various ML features that can access company data directly and create spreadsheets, graphs, and other visualizations.
The solution already has generative AI functionality with the integration of the AI assistant Amazon Q. This AI chatbot answers questions from end users, provides data summaries, and performs simple tasks based on normal language, such as generating dashboards.
Introduction scenarios
However, according to AWS, obtaining this information via a (traditional) AI chatbot often takes too long because users constantly have to adjust or customize their queries to get the best results.
In addition, until now, only ‘authors’ or creative professionals who build pages and visuals and consumers who want summaries and answers to questions could use the AI features in QuickSight.
With the introduction of the new agentic AI feature, also known as ‘scenarios’, more people within companies will have the opportunity to transform themselves into real data analysts without much prior knowledge, AWS says. This allows everyone to dive deep into all available data and get to work with spreadsheets and data sets.
The scenarios feature helps to speed up analyses and thus increase productivity by automating complex, multi-step analyses. In a scenario, users ask a question in plain English, after which the AI agent searches for it in the available data, formulates a plan and carries out the necessary (data analysis) steps. The tech giant indicates that this saves users time and a lot of effort.
During the AI agents’ activities, the entire analysis process is shown to the end user so that they can follow the path. This allows users to make adjustments because they can see how the AI agents thought and what choices were made to arrive at certain outcomes.
Adjusting is very easy
The AI agent’s entire “thought process” is visible in a thread, starting from the end user’s initial question. During the process, users can intervene by asking additional questions that may lead to more results or by considering other alternatives.
This enables them to avoid linear thinking when examining data and to explore multiple paths simultaneously. Ultimately, according to AWS, this would lead to more creative solutions.
The new agentic AI feature for Amazon QuickSight is now generally available.