Una

In this session I will be speaking about Kubernetes operator for AWS EKS that detects Spot interruptions on GPU nodes, triggers ML job checkpointing, and recovers training by rescheduling on available nodes. This I prepared with an idea to provide an interesting example of a repeatable workflow for maximizing GPU cost savings without losing progress on ML training jobs. The setup covered in this talk will include services AWS EventBridge, S3/EFS , IAM, and CloudWatch. I will be presenting a live example to show node interruption handling, and seamless ML job recovery using my developed solution.

Amazon Aurora DSQL is a serverless, distributed SQL database that brings together decades of Amazon innovation. In this session, explore how its active-active architecture delivers single- and multi-region resiliency with virtually unlimited scale. Learn key architectural principles, including transaction management and scaling, and get practical guidance for building applications on Aurora DSQL.

Embark on an interstellar journey through the lifecycle of a modern data platform! In this session, we’ll explore each evolutionary stage of a data platform, from humble beginnings to a fully-fledged, self-serve ecosystem. Drawing from real-world experience, I’ll share lessons learned—warts and all—on how and when to upgrade your platform as your organization’s data needs mature. Our voyage will chart the critical decision points: • Starting with early-stage analytics using AWS Athena, • Navigating challenges of scaling and complexity, • Upgrading with platforms like Snowflake and Redshift, • And eventually building a robust, self-service data solution that empowers your entire organization. Whether you’re just starting or seeking inspiration for your next leap, “The Hitchhiker’s Guide to the Data Platform” will offer field-tested guidance, practical AWS strategies, and a few memorable stories from the front lines. Don’t panic—grab your towel and get ready to demystify the universe of data platforms!

At this day and age where autonomous AI agents execute workflows to accomplish their objectives, event-driven architectures prove to be a very flexible way to integrate interactions between different agents and systems.

Formalizing their communication protocols and integration patterns is key to success not just today but in the years to come. In this talk, we will explore some practical approaches to this challenge.

Distributed systems unlock scalability beyond previous limits, yet they also introduce novel challenges and failure modes. Chaos Engineering advocates for the acceptance of these challenges as inevitable and promotes designing systems to accommodate them, rather than trying to combat them. With AWS Fault Injection Service (FIS) we prepare for these failure scenarios so they don’t catch us off guard. Live demos included🤞

We’ll present a comprehensive overview of the AWS technologies used at Bitstamp, along with alternative approaches, practical tips, and lessons learned. The talk will cover the full stack—from top to bottom—offering insight into how a crypto exchange leverages AWS to build and operate its infrastructure.

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

Practical implementations of vector databases for specific AI applications such as recommendation systems, semantic search, question-answering systems, and image recognition. Show how AWS vector database solutions can be used to create responsive, scalable AI applications