Warning: Creating default object from empty value in /home/customer/www/quarktech.com/public_html/wp-content/themes/salient/nectar/redux-framework/ReduxCore/inc/class.redux_filesystem.php on line 29
Natural Language Queries with QuickSight Q – Quark Technologies
Know-how

Natural Language Queries with QuickSight Q

By February 20, 2023 No Comments

Did you ever daydream about talking in person (or keyboard to input in this case) with a database and it returning exactly the results you want without getting into the intricacies of SQL or any other query language? Well, your wait is about to be over as AWS announced new capabilities of QuickSight during re:Invent 2022.

Amazon QuickSight Q, powered by machine learning, uses natural language processing to answer your business questions quickly. By using Q, you can save weeks of effort on the part of your business intelligence (BI) teams, who otherwise might have to build predefined data models and dashboards.

Q is optimized to understand business language that you use every day as part of your work, including phrases related to sales, marketing, and retail.
Unlike conventional natural-language query-based BI tools, Q uses machine learning to automatically understand the relationships across your data and build indexes. You can ask questions on all your data and get insights in seconds.

Q provides autocomplete suggestions, performs spell-checks, and suggests acronyms and synonyms that you can customize to be business-specific. Also, if Q gets an answer wrong, you can provide feedback to Q to correct the answer. This feedback is passed on to the BI team, who can fine-tune the data model or add more data.

Doing a quick recap, AWS introduced natural language queries to QuickSight in 2020 via QuickSight Q, making it possible for users to query data by typing plain English questions into a search bar such as “show me all users by country”. The latest announcements include allowing users to view business performance forecasts without involving any analysts or data scientists, by simply typing “give me the forecast” followed by different metrics and Q will deliver the appropriate visualizations for that query.

On top of that, Q also supports questions with “why” leading non-technical people into narrowing down the factors that led to a particular outcome from their data. This gives the possibility to start a query with “why” and then insert a metric and a timeframe such as:  “why did salaries drop in 2021?”

New features added to QuickSight include:
·        Paginated Reports, a tool for paginated reporting that provides summaries of data such as daily and weekly transactions
·        New APIs to help customers migrate data and analytics assets like reports and dashboards from on-premises databases
·        Improved SPICE with added power, double than previous support for datasets up to 1 billion rows

Future work on QuickSight includes improvements on data preparation and data governance with the aim of the improvements being targeted to three types of users: authors, readers and administrators.

As of now Q is already being used by organizations such as NFL, Capital One and Panasonic.

All rights reserved Salient.