We’ll kick-off our AWS Community Day with foreword from the organizers and a our Diamond Sponsors.

TBA

In today’s cloud-native world, treating infrastructure as code isn’t just a best practice—it’s a necessity. While YAML-based tools have served us well, they often fall short when dealing with complex, enterprise-scale infrastructure. The AWS Cloud Development Kit (CDK) bridges this gap by bringing software engineering principles to infrastructure deployment.

This session introduces AWS CDK as a powerful alternative to traditional IaC approaches, demonstrating how it enables teams to leverage familiar programming concepts while maintaining infrastructure consistency and reliability. We’ll explore how CDK’s programming model accelerates development cycles and reduces common infrastructure deployment pitfalls.

More and more companies are adopting serverless technologies as the community is defining the best practices, tools, and patterns.

Companies using Terraform as their infrastructure as a code tool are often required to reinvent the wheel when they work with serverless.

There is a misunderstanding, that everything is possible with the Serverless Services in AWS, for example that your Lambda function may scale without limitations. But each AWS service (not only Serverless) has a big list of quotas that everybody needs to understand and take into account during the development.

In this talk I’ll explain the most important things to be aware of for the scalability of the AWS Serverless services, explain quotas (from the hyper scalability point of view, but not only) of the services like API Gateway, Lambda, DynamoDB, Aurora Serverless, SQS, S3 and others and how to architect your solution with these quotas and technical concepts of the distributed systems in mind like token bucket algorithm, throughput and burst rate, concurrency, retry with exponential backoff and jitter and others.

2024 has been the year of taking generative AI applications into production. Launching generative AI applications requires careful considerations around model selection and evaluations, fine-tuning versus RAG, security, privacy, hallucination control, and cost management. In this session, dive into some common architecture patterns, security guardrails, governance approaches, and optimization tricks that have been developed to support hundreds of AWS customers launch their generative AI workloads in production globally across popular use cases, like content generation, chatbots, document search, and more.

Let’s explore some common costly mistakes in AWS and serverless and learn actionable tips for cutting down waste and reducing your AWS bill.

Whether you’re looking to cut down on CloudWatch costs or improve cost-efficiency for your serverless application, we’ve got some helpful tips for you.

AI isn’t replacing developers—it’s redefining their role. In this fast-paced 15-minute talk, we’ll explore how tools like AWS Bedrock for Agents are transforming coding into orchestration, and developers into architects and problem-solvers. Say goodbye to the 8-hour console grind and hello to workflows where AI handles the mundane.

Packed with real-world lessons (and a few funny AI fails), you’ll learn how to thrive in this new era—where the smartest developers aren’t writing more code, but guiding AI to do it for them.

DeepDoc Scanner is an AI-powered solution designed for searching structured and unstructured data efficiently. It automates document processing, eliminating manual work and minimizing the risk of errors. The system enables fast and precise retrieval of critical information, improving access to essential business insights. With built-in auditability and traceability, it enhances operational efficiency and compliance. DeepDoc Scanner supports digital transformation by converting paper-based documents into digital formats. It facilitates data categorization and analysis across multiple formats, including PDFs, Word, Excel, and OCR-processed images. The solution seamlessly integrates with existing data sources, ensuring smooth connectivity and accessibility. It features permission-based content filtering, allowing users to access only authorized information. DeepDoc Scanner includes various connectors and identity management capabilities to support secure data processing, making it a reliable and efficient tool for modern businesses.

This talk showcases how Stormit, as an AWS Advanced Partner, designed and delivered a scalable, event-driven data platform for a fast-growing biotech startup. Leveraging AWS services (S3, Glue, Athena, SQS, Lambda, Kinesis Data Firehose) and Snowflake’s Snowpipe, we automated data pipelines for multiple log types and analytics. Attendees will learn how to architect modern, serverless data pipelines, automate data ingestion and transformation.

Training an AI-powered Slackbot sounds straightforward - until your model starts ignoring half of the data you feed it. At AWS User Group Vienna, we built OTTO, a Slack-integrated AI assistant, fine-tuned using the open source tool InstructLab and deployed on Amazon Bedrock. But as we scaled up, we ran into real-world bottlenecks: training on MacBooks was slow, retrieval was inconsistent, and debugging was way harder than expected.

This talk goes beyond the ‘perfect AI stack’ and into the messy reality of model tuning, infrastructure choices, and the unexpected lessons we learned. If you’re working on AI-powered assistants (or just curious how fast things can go sideways), this talk will provide practical insights into our approach, with focus on cost efficiency, and what we learned on the way.

In today’s app development world, developers face many choices for handling authorization - from built-in framework solutions to third-party packages. Amazon Verified Permissions offers a cloud-based approach with its Cedar policy language, but is it worth adding to your tech stack?

Join this session for a hands-on look at AVP through practical examples. We’ll break down how Cedar’s policy language works, show you integration options that actually work, and honestly discuss the challenges you might face. Most importantly, we’ll dive deep into the pricing model to help you make an informed decision about whether AVP is the right fit for your authorization requirements.

As modern applications adopt cloud-native architectures, testing must evolve to keep up. In this presentation we’ll explore how Testkube, an open-source framework, streamlines testing workflows in Kubernetes environments.

We’ll discover how it integrates seamlessly with popular testing tools, how it fits naturally into CI/CD pipelines, and how its declarative configurations aligns with GitOps principles. Additionally, we’ll examine the benefits of centralized test reporting and how it helps to tackle scalability issues.

You’ve been hearing a lot about security best practices, but you’re not convinced they can really make a difference? Do you think your resources are safe only because nobody would notice your random IP address? In this presentation, I will share real-life attack scenarios to convince you that misconfigurations can have dire consequences.

Attendees will gain a comprehensive understanding of real-world examples, complete with screenshots. I will discuss multiple different attack vectors to demonstrate the diverse range of threats organizations face in today’s digital landscape.

This talk will challenge common misconceptions about AWS CloudWatch and highlight its often-overlooked capabilities.

I will provide a deep dive into CloudWatch, demonstrating why it deserves greater attention and consideration before committing to costly third-party observability solutions.

Let’s take a look at how to build Serverless apps in 2025 and what tooling to use. The talk will focus predominantly on AWS environment, but we might also take a peek at rising stars among the cloud providers like for example CloudFlare. We will take a look at SST.dev framework/tool which is built on top of Pulumi and focuses on providing the best Developer Experience when building cloud-native applications.

TBD

Event-Driven Architecture sounds great in theory, but how does it work in production? In this session, I’ll show you how we implement EventBridge in the real world, using practical examples from a blogging platform.

We’ll explore key architectural decisions like the subscription pattern and how to structure your events properly. Then we’ll get hands-on with TypeScript and CDK, covering event validation with zod and middy, error handling patterns, and implementing idempotency. All examples will be based on actual production code that you can adapt for your own projects.

When Every Millisecond Counts: Building a High-Performance Kubernetes Environment for a Multimedia Platform on AWS EKS Desc: Ever wondered what it takes to build a Kubernetes environment where every millisecond matters? This session dives into the architecture and optimizations behind a latency-sensitive multimedia platform running on AWS EKS. Discover how AWS Network Load Balancers, Ingress NGINX, and DNS optimizations like tweaking Node Local DNS and CoreDNS come together to deliver lightning-fast performance. Whether you’re dealing with similar challenges or just curious about Kubernetes optimizations, this talk has something for you.

Are you passionate about AI security and building resilient applications? This session delves into securing language models against prompt injection threats. Explore AWS deployment strategies, NLP input validation, prompt tracing, threat modeling, and effective countermeasures. Engage with practical insights and learn in a fun way.

Prompt injection is a new and increasingly serious threat to modern language models, jeopardizing the trustworthiness and security of AI-powered applications. But how can you protect applications built on these models? How can you properly design applications using LLMs? You’ll learn what prompt injection is, why it cannot be ignored, and how AWS helps create layers to protect language models. We’ll discuss proven strategies for developing applications resistant to prompt injection and input manipulation. Focus on NLP input validation. Additional countermeasures for security in LLM. Tracing of the prompts. Threat modeling of LLM. The presentation will be in the form of a storytelling or gamification experience.

Our Monitoring Journey: From Chaos to Nirvana

Alex Gain shares the evolution of his team’s monitoring setup — from manual chaos to a fully automated, GitOps-driven system managing dozens of Kubernetes clusters. The talk covers key transitions: early Terraform-based IaC, adoption of ArgoCD and ApplicationSets, and future plans with Crossplane. Packed with lessons on scaling observability, reducing engineer toil, and building resilient, client-agnostic monitoring platforms.

This talk explores approaches to managing and provisioning dynamic infrastructure for cell-based SaaS application, where each tenant requires dedicated resources due to security requirements, or gradual migration from silo based deployment.

Main focus areas include common resources for most SaaS applications, such as secrets backed by AWS Secrets Manager & KMS, database clusters backed by Amazon Aurora or S3 buckets. There will be examples and lessons learned throughout Ataccama’s year journey of creating a new SaaS, which is now in production.

Learn how to design and deploy autonomous, context-aware AI systems using agentic principles on AWS. Real-world projects demonstrate how business needs and technical requirements shape architecture and drive key decisions. By examining several customer-facing conversational systems, the session reveals common challenges, effective solutions, and valuable insights for moving from proof of concept to production-ready AI.

Closing ceremony of the AWS Community Day CZ 2025. We’ll wrap up the day with a few words from the organizers.