13:45 - 14:15 Session Block 5

In this session, we’ll explore how to design and deploy a scalable, serverless analytics pipeline on AWS that supports everything from business dashboards to advanced predictive insights. By leveraging a range of AWS services, we will demonstrate how to build a flexible data architecture that adheres to the AWS Well-Architected Framework. From data ingestion and transformation to data storage, querying, business intelligence, and advanced analytics, we’ll cover the entire process, ensuring a solution that can handle real-time and batch analytics while integrating predictive insights.

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🤞

At Loka we’ve planned and executed hundreds of migrations, and we know from experience that historically this very complex process can take months. But what if AI could evaluate your entire VMware estate, plan the optimal migration strategy, and refactor your applications–all before lunch? Our engineering teams are leveraging AWS Transform’s agentic AI capabilities to modernize legacy applications at scale and running 80% faster than traditional migration approaches. In this presentation, Loka’s Head of DevOps Miladin Papovic and Senior DevOps Lead Andrijana Stefanovska will pull back the curtain on Loka’s innovative migration process:

The evolution of generative AI has brought forth sophisticated multi-agent systems (MAS) that can tackle complex tasks through distributed intelligence and collaborative problem-solving. These systems represent a paradigm shift from traditional single-agent AI implementations to coordinated networks of specialized agents working in concert. As companies continue their digital transformation journeys, understanding the fundamental patterns of multi-agent architectures becomes crucial for designing robust, efficient, and scalable AI solutions. This workshop examines the core patterns in multi-agent systems—Parallel, Sequential, Loop, Router, Aggregator, Network, and Hierarchical—with specific applications in the financial services domain. We’ll explore how these architectures can revolutionize workloads across financial services through enhanced reasoning, planning, and collaborative intelligence. Expected outcomes:

Orbit, the backbone of BMW’s Connected Drive, is a dynamic ecosystem with over 360 EKS clusters supporting more than 22 million connected vehicles. As part of our continuous improvement, Orbit is transitioning from Cluster Autoscaler to Karpenter, marking a significant step towards operational excellence and cost-effectiveness. We will discuss how Karpenter’s features improved resource utilization, have reduced costs and streamlined operations.Additionally, we will highlight the enabling capabilities of ARM nodes within Karpenter, focusing on the advantages of utilizing ARM chipsets in our infrastructure. Our rollout strategy has ensured a seamless migration while maintaining the reliability essential for Connected Drive’s mission-critical services. This session will provide insights into the operational and financial benefits of Karpenter and ARM nodes, sharing our experiences, best practices, and the measurable impact on our infrastructure’s performance and sustainability.