I am still buzzing from attending the keynotes from Amazon Web Services CEO Adam Selipsky on Day 2 and VP of Analytics, Databases and Machine Learning Dr Swami Sivasubramanianon Day 3 who shared new announcements for Amazon Q.
As a data scientist, data analyst, developer you may be spending up to an hour each day writing production code to integrate and transform multiple datasets, build a data model and then design interactive dashboards to uncover business insights.
Let me introduce you to Amazon Q and we will learn how to incorporate generative AI to create dashboards using natural language processing for supercharging data storytelling.
In this lesson you will learn the following:
- What is Amazon Q?
- What is Amazon QuickSight?
- What are the benefits?
- What are the features?
- How do I get started?
- Build a dashboard using Amazon Q in Amazon QuickSight(Preview)
Amazon Q is an AI-powered assistant designed for work and can be tailored to your business by taking actions, answering questions, generating content and helping you to solve problems.
Amazon Q in QuickSight is now available in public preview which means that you can design dashboards using natural language processing by using a free-trial period of 30 days and you will not be charged under the two available pricing plans i.e. Amazon Q Business or Amazon Q Builder which is coming soon.
You may find out more from this video below, Introducing Amazon Q
Amazon QuickSight is a cloud-powered business intelligence service that allows you to build interactive dashboards that includes machine learning insights that provides enterprise security and can easily scale.
You may connect to data from multiple sources including AWS services, spreadsheets and third-party databases.
As a data scientist, data analyst or business analyst you will not need to code in any programming language with Amazon Q in Amazon QuickSight.
Some key benefits include:
Reduce time spent to build interactive dashboards to uncover data insights
Ask specific questions using natural language e.g. drill-down or filter data
Amazon Q caters to data and security needs with security in the cloud under the AWS Shared Responsibility Model by understanding your governance requirements for roles and permission levels.
Create visually appealing and compelling narratives
Amazon Q in QuickSight (Preview) includes features that allow you to speed up the delivery of data insights to your business stakeholders by using natural language processing.
The summary of key features include:
Obtain key highlights by asking Amazon Q to build an executive summary which takes a few seconds by leveraging large language models.
Produce contextual narratives from questions asked using Amazon Q.
Amazon Q allows you to drill-down into information on the dashboard.
data Q&A Question and Answering capability of Amazon Q allows you to dig deeper into your data by asking questions beyond your dashboard.
You may get started with Amazon Q in QuickSight by accessing the free trial in Preview.
If you would like to find out more details, you may watch the session on how to get started with Amazon Q from VP of Amazon Matt Wood at AWS re:Invent 2023:
Let's gets hands on practice to build a dashboard powered by generative AI.
- Step 1: Find an interesting dataset that is open-source.
I have selected the Starbucks from Kaggle.com
Step 2: Sign into the AWS Management Console as an IAM user
Step 3: In the search bar type 'Amazon Q'.
Switch to a supported AWS region e.g. US East(N.Virginia)
- Step 4: Click Free Trial for Amazon Q in QuickSight
You may select Enterprise + Q which is a 30 day free trial for your team.
Click Continue and enter your contact details.
- Step 5: Create your QuickSight Account and under pricing click No. Maybe Later.
- Step 6: Select an appropriate Authentication method.
I have an existing AWS IAM credential and I will select the first option.
Secondly, create a QuickSight Account Name and enter an email address.
- Step 7: Create a folder in your Amazon S3 bucket and upload the Starbucks csv file.
- Step 8: Select the Amazon S3 bucket to access the data.
- Step 9: You will receive a notification message that your Amazon QuickSight account has been created.
- Step 10: You will be directed to the Amazon QuickSight welcome page of your account.
- Step 11: Click New Analysis.
- Step 12: Select New dataset.
- Step 13: Connect to data by selecting the option Upload a file and select the Starbucks csv file saved in your local directory.
- Step 14: Confirm file upload settings and click Next.
- Step 15: Click Edit Settings and Prepare Data to inspect the data and check the correct data types.
- Step 16: Click Publish and Visualize.
Click Build Visual and create a topic to ask questions using Amazon QuickSight.
I have named the topic 'Coffee Details' and click Link Topic.
Amazon Q is preparing the topic in QuickSight.
- Step 17:The topic 'coffee details' has successfully been created and we will navigate to the Q bar.
- Step 18: We will start by asking a few questions in Q bar.
Amazon Q has recommended a few suggested questions to ask before I start typing the business question.
- Step 19: I click on 'Coffee Details', there is no narrative provided as this is not a question.
- Step 20: I click Ask Q and select a suggested question:
'total cholesterol mg by beverage?'
- Step 21: I click Ask Q and select a suggested question:
'Top 5 beverages by total calories'
- Step 22: I click Ask Q and select a suggested question:
'Beverage category with the highest total fat by category'
- Step 23: Click Ask Q to select a suggested question:
'Beverage category with the highest total sugar by category'
- Step 24: Navigate to Named Entity and group variables.
I have grouped trans fat, saturated fat and total fat.
- Step 25: I have created a second named entity for 'Vitamins'.
- Step 26: I have created a third named entity for 'Sugar' by entering at least two dimensions and clicking Save.
- Step 27: I have created a fourth named entity for 'Nutrition value' by entering at least two dimensions and clicking Save.
- Step 28: We can navigate back to Ask Q such as the total amount of sugar in each beverage and click the light bulb icon to view insights.
- Step 28: Five visuals have been saved as pinboard.
To build a dashboard, you may select Add to analysis and choose four pinboard visuals created by Amazon Q.
You may then check the boxes for Allow executive summary and publish the dashboard.
- Step 29: Let's create a data story.
Navigate to Data stories on the left-handside and select New story. Select Slideshow.
Enter a 'Story title' and describe the story in simple natural language by entering a prompt and Add Visuals.
Amazon Q was able to generate content via Data stories to create a slide presentation deck based on my visuals with AI-powered data insights.
- Step 30: By navigating to my dashboard, on the right-handside under Build select from the drop-down menu Executive Summary and click on Sheet 1.
In a few seconds, content is generated for an Executive Summary as shown below:
You may easily create a dashboard using generative AI to supercharge data storytelling with large language models using Amazon Bedrock to generate insights, suggested questions and even help you create a slide deck with data stories.
An executive summary is content created from your visuals and natural language prompt that you provide on a selected worksheet.
It is very easy to get started under Preview with the 30 day free-trial.
Until the next lesson happy learning! 😁
You may watch this video to learn how to use Generative BI with Amazon Q to build dashboards in Amazon QuickSight:
If you would like to learn more, be sure to have a look at Amazon QuickSight Community:
You may watch these highlight sessions discussing Amazon Q from AWS re:Invent 2023 last week from Las Vegas. I recommend the following to watch:
- Senior Principal Engineer, Clare Liguori - Amazon Q: Your new assistant and expert guide for building on AWS
- CEO at Amazon Web Services, Adam Selipsky's Day 2 keynote, watch Amazon Q announcements from timing [01:22]
You may find more details for Re-cap sessions happening this week in your local city.
If you are located in Sydney, you may register to attend AWS re:Invent 2023 recap on Thursday 14 December this link