My Takeouts from AWS Cloud Day Prague 2023

A few days after attending the AWS Cloud Day Prague event that took place at Forum Karlín on October 24, 2023, I had a time to reflect on it and decided to write about topics that I found most interesting and fun. This article is very subjective, and I don’t aspire to cover 30+ breakout sessions that took place that day, nevertheless hope that some of the information might be useful for readers that did not attend the event or its alternative in other cities.

Right after I entered the event, I saw the eye-catching serverlesspresso, a system for coffee ordering powered by AWS serverless technologies. The machine was connected to various AWS serverless services to demonstrate how they can work together to create a scalable and reliable serverless application. It can process a thousand coffee orders with an AWS bill of around $9. The attendees could order their coffee by scanning a QR code and monitor the status of their order and the machine’s performance through a web dashboard. This idea appeared before on re:Invent 2021 and you can see more details about at Learn more about Serverlesspresso! | Serverless Land.

The star of the event was definitely artificial intelligence and machine learning. I am very glad to see that AWS did not drop the ball and is pushing hard to be a relevant provider when it comes to the AI/ML solutions. Many AI/ML sessions took place that day, accompanied by AWS booths dedicated to AI/ML applications and services.

I enjoyed sessions shedding light on the rich SageMaker ecosystem for managed machine learning services that simplifies the entire machine learning lifecycle. It offers:

• SageMaker JumpStart: Easy access to pre-built solutions, models, and notebooks for common machine learning use cases.
• SageMaker Wrangler: Simplified data preparation for machine learning.
• SageMaker Autopilot: Automated machine learning tasks such as data preprocessing, algorithm selection, hyperparameter tuning, and model deployment.
• SageMaker Feature Store: Feature creation, storage, sharing, and access for machine learning.
• SageMaker Pipelines: CI/CD service for machine learning workflows.
• SageMaker Studio: Web-based IDE for machine learning.

There are plenty of SageMaker workshops available at AWS Workshops, such as Low-Code ML with SageMaker for Data Scientists (workshops.aws) and Amazon SageMaker Data Wrangler — Deep Dive Workshop (workshops.aws).

The second big AI/ML topic of the event was dedicated to Amazon Bedrock, AWS’s service for building and scaling generative AI applications using foundation models. I realized how varietal options are there for RAG-based search, depending which services you have knowledge or preference working with. Whether it comes to the large language model choosing (Titan, Claude, Llama, …) or contextual embedding store (Kendra, OpenSearch, MongoDB, ...), there is plenty of flexibility in AWS to compose architecture that suits your needs.

You can take a look at GitHub repository aws-samples/amazon-bedrock-workshop: This is a workshop designed for Amazon Bedrock a foundational model service. (github.com) and Building with Amazon Bedrock and LangChain (workshops.aws) to get started with Bedrock and get to know it better.

Lastly, I’d like to highlight Connected Foosball Table, where our group spent time when we wanted to relax a bit. I was impressed by the clever combination of AWS services to achieve real-time capturing of the foosball playground to calculate game metrics. This originally appeared on AWS Summit Paris 2023, and you can find more information about it at Connected Foosball Table | LinkedIn. We were so inspired by it that we plan to implement the idea on top of other game!

AWS Cloud Day Prague was a great experience for me, and I learned a lot from it. I want to thank AWS for organizing this event, and the speakers and the participants for sharing their knowledge and ideas. I would love to hear your thoughts on the event, or the topics discussed, so feel free to leave a comment below or contact me via email or social media.

We are ACTUM Digital and this piece was written by Milan Gatyás, Cloud Architect & Team Lead of Apollo Division. Feel free to get in touch.