Build an assistant to AI of a multiple interface with Amazon Q and Slack with Amazon Cloudfront References clickable of a Amazon Cube S3

There are consistent comments of customers that IA attendees are the most useful when users can interact with them within the productivity tools they already use daily, to avoid changing applications and context. Web applications such as amazon Q Business and Loose They have become essential environments for the deployment of modern the assistant. This publication explores how the various interfaces improve user interaction, improve accessibility and serve different preferences.

By offering perfect experiences in environments, organizations can increase user satisfaction and adoption rates. The assistant uses the augmented generation of recovery (RAG), a technique that integrates credible and authorized sources within the answers between these interfaces, reinforcing reliability and educational value. This multiple interface approach and with rag impulses not only strives to meet the flexibility demands of modern users, but also encourages a more informed and compromised user base, finally maximizing the effectiveness and scope of the assistant. When combining RAG with multiple interfaces, the assistant offers consistent, precise and contextually relevant information, regardless of the user's favorite environment and productivity tools.

General solution of the solution

The following diagram illustrates the architectural design of the application.

<img class="aligncenter wp-image-97506 size-full" style="margin: 10px 0px 10px 0px;border: 1px solid #CCCCCC" src="https://d2908q01vomqb2.cloudfront.net/f1f836cb4ea6efb2a0b1b99f41ad8b103eff4b59/2025/01/18/ML-17585-Build-a-multi-interface-ai-assistant-using-amazon-Q-and-Slack-with-amazon-CloudFront-clickable-references-from-an-amazon-S3-bucket-architecture.png” alt=”Build an assistant to the multiple interface using amazon Q and Slack with amazon Cloudfront References clickable from a amazon Cube architecture S3″ width=”936″ height=”542″/>

You can find the complete code and the steps to implement the solution in the <a target="_blank" href="https://github.com/aws-samples/Multi-Interface-Chatbot-using-amazon-Q-and-Slack-with-CloudFront-Clickable-References/tree/main” target=”_blank” rel=”noopener”>Github repository.

Click here to open the AWS and continue console.

Previous requirements

You must have the following previous requirements:

Implement the solution

For configuration steps, see the <a target="_blank" href="https://github.com/aws-samples/Multi-Interface-Chatbot-using-amazon-Q-and-Slack-with-CloudFront-Clickable-References/blob/main/README.md” target=”_blank” rel=”noopener”>Readme In the Github repository.

Solution components

In this section, we discuss two key components of the solution: the data sources and the vector database.

Data sources

We use Spooky documentation RST files (restructured text) loaded in a amazon simple storage service cube (amazon S3). Every time the assistant returns it as a source, it will be a link in the specific part of Spack's documentation and not The upper part of a page of origin. For example, Spack images in Docker Hub.

Creepy It is a versatile package administrator for supercomputers, Linux and macOS that revolutionizes the installation of scientific software by allowing multiple versions, configurations, environments and compilers to coexist in a single machine. Developed by Todd Gamblin in the National Lawrence Livermore Laboratory In 2013, Spack addresses the limitations of traditional packages administrators in high performance computer science (HPC) environments. Brian Weston, Cloud for Mission Science transformation program in LLNL, advised in the development of this assistant.

In addition, we use text files loaded in a s3 cube that can be accessed through an amazon Cloudfront link. There is also an automated ingestion work from Slack conversation data to Cube S3 driven by an AWS Lambda function. This allows the assistant to also use previous conversations of users to answer questions and quote their sources. We chose to use Cloudfront links instead of using Slack links because when this source is cited on amazon Q, the user might not have access to Slack data. There is also an alternative to this methodology using the Slack connector for amazon Kendra.

This solution could admit other types of data, such as PDF, Word documents and more, as long as your text can be extracted and feed on the vector database with some changes in the code. Your RAW files can be served in a cloudfront distribution.

The next screen capture illustrates a sample cloud sample url.

<img loading="lazy" class="aligncenter wp-image-97507 size-full" style="margin: 10px 0px 10px 0px;border: 1px solid #CCCCCC" src="https://d2908q01vomqb2.cloudfront.net/f1f836cb4ea6efb2a0b1b99f41ad8b103eff4b59/2025/01/18/ML-17585-Build-a-multi-interface-ai-assistant-using-amazon-Q-and-Slack-with-amazon-CloudFront-clickable-references-from-an-amazon-S3-bucket-cloudfront.png” alt=”Build an assistant of a multiple interface with amazon Q and Slack with amazon Cloudfront References Clickable from a amazon S3 Bucket Cloudfront” width=”936″ height=”198″/>

After the implementation, the existing data is automatically charged in a s3 cube and are processed to be used by the assistant. The solution also includes Slack's automatic daily data ingestion using amazon Eventbridge.

Vector database

This solution uses amazon Kendra as its vector database, offering significant advantages in simplicity and profitability. As a fully managed AWS service, amazon Kendra reduces development and maintenance costs. amazon Q, which admits two types of retrievers (Native Retriever and amazon Kendra), integrates perfectly into this configuration. When using amazon Kendra, the solution efficiently uses the same retriever for both amazon Q and Slack interfaces. This approach not only speeds up general architecture, but also provides a user experience more consistent in both environments. The result is a cohesive and profitable system that maintains uniformity in the recovery and presentation of the information, regardless of the interface chosen by the user.

amazon Kendra also admits the use of metadata for each archive file, which allows both IU to provide a link to its sources, whether the Spack documentation website or a Cloudfront link. In addition, amazon Kendra admits a relevance adjustment, which allows us to increase certain data sources. For this solution, we increase the results for Spack's documentation.

User interfaces

In this section, we discuss the UIs used in this solution.

amazon Q Business

amazon Q Business uses RAG to offer an insurance assistant and knowledge of knowledge adapted to your organization. As AWS native solution, it integrates perfectly with other AWS services and presents its own easy -to -use interface. This integration, combined with its simple configuration and implementation process, provides an implementation experience without problems. By merging the generative capacities of ai with the recovery of intelligent information of their business systems, amazon that Business offers precise and aware of the context firmly rooted in the specific data and documents of its organization, improving its relevance and precision.

The next screen capture is an example of the amazon Q Business user interface.

<img loading="lazy" class="aligncenter wp-image-97505 size-full" style="margin: 10px 0px 10px 0px;border: 1px solid #CCCCCC" src="https://d2908q01vomqb2.cloudfront.net/f1f836cb4ea6efb2a0b1b99f41ad8b103eff4b59/2025/01/18/ML-17585-Build-a-multi-interface-ai-assistant-using-amazon-Q-and-Slack-with-amazon-CloudFront-clickable-references-from-an-amazon-S3-bucket-amazon-q.png” alt=”Build an assistant to ai of a multiple interface with amazon Q and Slack with amazon Cloudfront References by clicking on a amazon S3 Bucket amazon Q” width=”936″ height=”502″/>

Loose

Loose It is a popular collaboration service that has become an integral part of the communication forums of many organizations. Its versatility extends beyond equipment messages to serve as an effective interface for attendees. By integrating attendees with ai in Slack, companies can use their family environment to provide users with instant access to information.

The next screen capture shows an example of the Slack user interface with a message thread.

<img loading="lazy" class="aligncenter wp-image-97510 size-full" style="margin: 10px 0px 10px 0px;border: 1px solid #CCCCCC" src="https://d2908q01vomqb2.cloudfront.net/f1f836cb4ea6efb2a0b1b99f41ad8b103eff4b59/2025/01/18/ML-17585-Build-a-multi-interface-ai-assistant-using-amazon-Q-and-Slack-with-amazon-CloudFront-clickable-references-from-an-amazon-S3-bucket-Slack.png” alt=”Build an assistant of a multiple interface using amazon Q and Slack with amazon Cloudfront References Clickable from amazon S3 Bucket Slack” width=”936″ height=”666″/>

Listen

amazon q has a built -in feature for an analysis panel that provides information on user participation within a specific environment of amazon Qat Consult, which allows you to analyze and optimize the performance and interaction of the user of your IA assistant.

For Slack, we are collecting comments from users, as shown in the previous screenshot of the user interface. Users can add a “thumb up” or a “thumb down” to the response of the assistant to monitor their performance. In addition, we have created a personalized solution that uses a Cloudwatch amazon board to imitate the amazon analysis board Q to further align the experience between the two applications.

The next screen capture shows an example of the Slack Cloudwatch board.

<img loading="lazy" class="aligncenter wp-image-97508 size-full" style="margin: 10px 0px 10px 0px;border: 1px solid #CCCCCC" src="https://d2908q01vomqb2.cloudfront.net/f1f836cb4ea6efb2a0b1b99f41ad8b103eff4b59/2025/01/18/ML-17585-Build-a-multi-interface-ai-assistant-using-amazon-Q-and-Slack-with-amazon-CloudFront-clickable-references-from-an-amazon-S3-bucket-cloudwatch.png” alt=”Build an assistant to the a multiple interface using amazon Q and Slack with amazon Cloudfront References by clicking on a amazon S3 Bucket Cloudwatch” width=”936″ height=”580″/>

In addition, there is a daily scheduled slack message that summarizes Slackbot data during the last day, as shown in the next screenshot.

<img loading="lazy" class="alignnone wp-image-97836 size-full" style="margin: 10px 0px 10px 0px;border: 1px solid #CCCCCC" src="https://d2908q01vomqb2.cloudfront.net/f1f836cb4ea6efb2a0b1b99f41ad8b103eff4b59/2025/01/23/ML-17585-Build-a-multi-interface-ai-assistant-using-amazon-Q-and-Slack-with-amazon-CloudFront-clickable-references-from-an-amazon-S3-bucket-reports-resized.png” alt=”Build an assistant of a multiple interface with amazon Q and Slack with amazon Cloudfront References Clickable from a amazon S3 Bucket Reports” width=”500″ height=”250″/>

Clean

To avoid incurring ongoing positions, clean the resources it created as part of this publication with the command mentioned in the <a target="_blank" href="https://github.com/aws-samples/Multi-Interface-Chatbot-using-amazon-Q-and-Slack-with-CloudFront-Clickable-References/blob/main/README.md#clean-up” target=”_blank” rel=”noopener”>Readme.

Conclusion

The implementation of a multi -interface IA assistant using RAG represents a leap in organizational communication promoted by ai. By integrating amazon Q Business and Slack interfaces with a solid backend -driven backend Kendra, this solution offers perfect and agnostic access to the environment to precise and conscious information of the context. The strengths of architecture are found in their consistency in environments, automatic data ingestion processes and comprehensive monitoring capabilities. This approach not only improves user participation and productivity, but also positions organizations to quickly adapt to evolving communication needs in an increasingly centered panorama, which marks a fundamental step towards management systems of more efficient and intelligent information.

For more information about AWS services used in this solution, see the amazon Q user guide, implements a Slack bonding door for amazon Bedrock and the amazon Kendra developer guide.

About the authors

Nick Bison He is an automatic learning engineer at Aws Professional Services. Solve complex organizational and technical challenges using data and engineering science. In addition, it builds and implements ai/mL models in the AWS cloud. His passion extends to his propensity to travel and various cultural experiences.

Dr. Ian Linsford He is an aerospace consultant in the cloud at AWS Professional Services. Integrate cloud services into aerospace applications. In addition, Ian focuses on building ai/mL solutions using AWS services.