Fixing Alert Fatigue: Modern Azure Monitoring Designs

Setting arbitrary metric thresholds like "Alert if CPU > 80%" is a broken approach to cloud operations that floods engineering teams with noise while missing actual system failures. Modern, high-concurrency cloud ecosystems require a shift from reactive monitoring to proactive observability. This purely conceptual session breaks down the architectural patterns needed to design an intelligent, high-signal alerting engine on Microsoft Azure.

We will analyze structural blueprints for centralized telemetry collection using Azure Monitor and Log Analytics workspaces. We'll explore how to design dynamic, anomaly-based alerting models, establish proper alert routing paths, and structure multi-tiered notification systems that separate background noise from critical infrastructure failures.

Architecting Backend Infrastructure for Autonomous AI

Deploying AI agents into an enterprise environment requires far more than just basic API wrappers; it demands an underlying infrastructure that can handle long-running, asynchronous agent tasks and unpredictable state orchestration. This technical session explores the backend patterns required to build a resilient, scalable home for autonomous AI workflows.

We will unpack architectural frameworks for managing agentic state machines, handling high-concurrency requests without starving core services, and setting up secure operational governance. Attendees will walk away with an enterprise-ready infrastructure blueprint built to scale agentic workloads while maintaining tight control over cloud resource utilization.

Event-Driven Autoscaling on AKS Workflows using KEDA

Moving complex enterprise platforms to production requires robust, resilient cloud-native infrastructure. However, traditional resource-based autoscaling (such as standard CPU or Memory metrics) fails to react quickly enough to unpredictable traffic spikes and high-concurrency data streams passing through distributed backend layers.

In this deeply technical session, we will break down how to architect a highly responsive backend system using Azure Kubernetes Service (AKS). We will explore real-world production blueprints for implementing Kubernetes Event-driven Autoscaling (KEDA) to scale microservices dynamically based on actual application demand, system queue depths, and message brokers.

Architecting Tech Startups: Tech for Sustainability

For tech startups launching global sustainability platforms, the initial engineering challenge is managing massive, disparate datasets across enterprise supply chains. Building a platform from the ground up requires a robust cloud architecture capable of handling high-concurrency data logging.

Using the engineering journey of building Raggedy Bag (raggedybag.com)—an open-beta enterprise platform—this talk details how to establish distributed backend systems and optimized cloud infrastructure to turn complex tracking into real-time intelligence.

This session provides tech founders and software architects with a practical blueprint for deploying resource-smart orchestration layers to solve critical B2B sector challenges at speed.

From Code to Company: The Tech Startup Founder Journey

Transitioning from an engineer to a technical founder requires a massive shift in mindset, motivation, and execution. Building and launching an open-beta B2B platform involves navigating unique hurdles—from managing initial resource constraints to designing infrastructure that can survive day-one enterprise demands.

This session shares the raw realities, critical milestones, and core motivations behind building a tech startup from scratch. Tech founders and developers will walk away with a practical blueprint for balancing technical architecture with strategic business growth.

Target Audience: Developers, tech leaders, and aspiring founders.

Presented by Star Lab Systems — Innovate. Integrate. Inspire.
https://starlabsystems.uk/

Building Better Developer Paths: Infra as a Product

Developers shouldn't have to be cloud infrastructure experts to deploy code securely. This session unpacks how to treat your internal developer platform as a product, creating "golden paths" that abstract away Kubernetes complexity. Learn how to design predictable, self-service infrastructure blueprints that allow engineering teams to move quickly without risking production drift or security vulnerabilities.

Designing for Failure: Microservices on Distributed Cloud

In a high-concurrency B2B ecosystem, network drops and downstream dependency failures are inevitable. This deeply technical session explores architectural patterns, such as circuit breakers, retry backoffs, and asynchronous message decoupling ,needed to ensure your services remain resilient. We will map out how to build fault-tolerant backend architectures that prevent localized outages from collapsing your entire cluster.

Lean Engineering: Prioritizing Tech Debt in Startups

Early-stage technical founders constantly wrestle with the trade-off between shipping features fast and writing perfectly scalable code. When does "good enough" stop working? This session shares an objective framework for managing technical debt while keeping your startup lean. We will outline how to audit your architecture, identify which bottlenecks will actually break your next milestone, and intentionally phase out shortcuts as your platform scales.