SREDAY

When you're managing dozens of Kubernetes clusters, across three different clouds, for dozens of individual companies in their own accounts, the challenge of (re)designing tooling is complex. Come hear how we worked through all the many possible options (centralized IaC vs templating, whether to use cluster managers like Rancher, etc.), what high-level tradeoffs were made (i.e. ease-of-use vs speed of changes vs centralized control), and the tangible outcomes of this process. The focus here will be on the process and decision making, given all the different drivers such as technology, business, people, process, etc. This talk will cover how we undertook the process, while giving you tangible next steps to take back to your desk.

From this talk, you'll take away: - Some reasons why you should or should not attempt a rewrite of your Infrastructure-as-Code process. - An overview of the inputs that you should consider when designing an Infrastructure-as-Code tooling stack. - Some idea of how to be successful with a tooling rewrite.

Andrew Suderman has had a lifelong obsession with technology that currently fuels his passion for Kubernetes. He enjoys designing and building new systems that make life easier for developers. Currently he pursues that passion at ReactiveOps, where he maintains dozens of clusters and works on internal tooling.

In his spare time he enjoys off-roading, camping, quadcopters, motorcycles, and far too many other hobbies to afford.

Zero Trust doesn't have to be intimidating. Learn how Identity-Aware Proxies transform service access from perimeter-based to continuous verification, explained through the universal experience of airport security.

Imagine if showing your ID once at airport security meant you could then board any plane, enter the cockpit, or walk into customs. Sounds absurd? Yet this is how perimeter-based security works - one check, then complete trust. Based on Google's BeyondCorp papers, this talk breaks down Zero Trust and Identity-Aware Proxies, where every access request is verified based on identity, device, and context - just as airports verify your credentials at every secure checkpoint, from airport security to boarding gates to airline lounges. Through technical concepts and the analogy of airport security, we'll explore how to implement these principles in real-world applications.

Nick is a developer advocate at Pomerium, a zero trust, identity-aware proxy platform that enables secure, clientless connections to web applications and services without a corporate VPN.

With over a decade of open source contributions and five years of professional open source experience at companies like OpenSauced, dev.to, and Netlify, he brings deep community knowledge to his work. You’ll often find him live streaming tech content, either solo or with friends from the community.

In today's fast-paced world, Agile & security is paramount. Urgent Cloud Services Deployment Practices ensure seamless and secure cloud deployments under tight timelines. Gain actionable insights, maintain robust security, and enhance team coordination for efficient, reliable cloud infrastructure.

Sangeetha Rajkumar is a seasoned technologist and product manager with over 9 years of experience in software product development. As a Senior Technical Program Manager at Microsoft, she leads the development of innovative cloud products for Microsoft customers. Sangeetha has spearheaded numerous automation projects in Deployment and DevOps, showcasing her technical expertise, strong product & program management skills.

She holds a master's degree in engineering management with a specialization in Product Management from Cornell University. With nearly a decade of experience in the IT industry, Sangeetha has worked across various sectors, including Consulting, Fintech, and Enterprise Solutions. Additionally, she mentors early-career professionals at Microsoft, ADPList, and Topmate.io.

Learn how to identify hidden patterns in cluster behavior that impact reliability. Discover methods to detect issues like pod delays and service dependencies, using clustering algorithms and proactive strategies for improved resilience and resource optimization.

Join us for a deep dive session that extracts data-driven observations and insights from heavily utilized clusters to find unique patterns that impact reliability. We'll share how to implement the right signals to detect common challenges such as delays in pod spin-ups, pod termination data to identify workloads most affected by these terminations, or closely occurring failures that indicate hidden service dependencies. We'll then cover methods to preemptively address these to improve resilience using clustering algorithms like DBSCAN, for dependency mapping, and other good practices. We'll walk through a typical journey from experimental setups using scripts and command-line outputs to production-ready practices––from quick prototyping, to refining data pipelines and optimizing performance for operational environments. By uncovering these hidden patterns & signals you can achieve proactive reliability improvements, for improved cluster health & optimization of resource allocation.

Gary Edgar is a seasoned technical support and customer engineer with 15+ years of experience across cloud infrastructure, databases, and monitoring systems. Currently at Komodor, he helps teams streamline Kubernetes troubleshooting.

Previously, Gary was a key early hire at PlanetScale, supporting major clients like Stripe and Twitter with performance tuning, architecture, and hands-on support. He also delivered impactful solutions during his time at SevOne (IBM), working onsite with Microsoft and eBay to improve system reliability and monitoring at scale.

For years, you've hopefully been relying on the Microsoft Well-Architected Framework (WAF) as a starting point to have your workloads set up in a highly available, redundant architecture. Following those guidelines is key to get reliably running business critical applications for your users or customers.

But even after following all the guidelines, did you really know your applications would run as expected? With this session, I'm going to take the guesswork out of that question, and walk you through different tools in Azure, Azure Load Testing and Azure Chaos Studio, which help by stress-testing your critical workloads, running in App Services, Containers, CosmosDB or traditional Virtual Machines.

As usual in Peter's sessions, this will be packed with live demos and only a bare minimum of slides.

Peter has an extensive background in architecting, deploying, managing and training on Microsoft technologies, dating back to Windows NT4 Server in 1996, all the way to the latest and modern cloud solutions available in Azure today. With a passion for cloud Architecture, Devops and App Modernization, Peter always has a story to share on how to optimize your Enterprise-ready cloud workloads. When he's not providing a technical Azure workshop out of his role as Microsoft Technical Trainer at Microsoft Corp, for which he relocated from Belgium to Seattle early 2022, he's developing some new apps in .NET Blazor, or trying another bread baking recipe in his home kitchen. Peter was an Azure MVP for 5 years, is a Microsoft Certified Trainer (MCT) for +15 years and still actively involved in Tech Communities as public speaker, keynote presenter, technical writer and mentor.

This talk is for DevOps and SRE engineers who are looking to optimize the number of SLOs in their system and find methods to manage a large number of various SLIs for their complex, multi-level service architecture. The talk covers practical methods for calculating SLOs based on multiple SLIs and dependencies, including the reliability of third-party service providers and unknowns. Attendees are expected to be familiar with basic SRE principles and terminology. The talk includes some probability math.

Overwhelmed by the complexity of managing SLOs across your microservices architecture? This talk provides DevOps and SRE engineers with actionable strategies for defining, calculating, and optimizing SLOs in systems with a large number of SLIs.

We’ll explore practical techniques for:

• SLO Targets: Right-guessing and improving your SLO goals.
• SLO Optimization: Right-sizing the number of SLOs to ensure effective monitoring without unnecessary overhead.
• SLO Composition: Mastering aggregation and other methods to derive meaningful SLOs from diverse SLIs.
• Dependency Management: Incorporating the reliability of third-party services and unknowns into your SLO calculations.

The talk will include some probability math to illustrate key concepts.

In my work at Google as a Developer Relations Engineer, I help cloud practitioners to solve a wide range of technical and architectural problems related to running workloads in Google Cloud and hybrid environments. Jump start to Google Cloud. Reach 99.99 SLO. Troubleshoot a problem in real time. Or, evaluate a security posture of your application. All these and many other problems that engineers and architects meet every day. Before Google, I worked in healthcare, interactive TV and gaming industries in many engineering roles including SWE, TL and Cloud Architect. In my free time I enjoy swimming, hiking and horseback riding.

Maintaining robust network security in the cloud environment is more crucial than ever. This talk explores the design and implementation of a proactive network telemetry platform that strengthens infrastructure security. The platform collects and analyzes telemetry data from virtual machines to identify and scrutinize traffic patterns deviating from industry-standard baselines. By generating actionable insights and promptly notifying relevant teams of potential issues, it enables swift problem resolution and fosters collaboration among development, security, and operations teams. This proactive DevSecOps approach significantly reduces the attack surface, minimizes exposure to network attacks, and enhances overall cloud security.

Siri Varma Vegiraju is a seasoned professional in healthcare, cloud computing, and security. Currently, he focuses on securing Azure Cloud workloads, leveraging his extensive experience in distributed systems and real-time streaming solutions. Prior to his current role, Siri contributed significantly to cloud observability platforms and multi-cloud environments. He has demonstrated his expertise through notable achievements in various competitive events and as a judge and technical reviewer for leading publications. Siri frequently speaks at industry conferences on topics related to Cloud and Security and holds a Masters Degree from University of Texas, Arlington with a specialization in Computer Science.

The adoption of AI-powered procurement is revolutionizing supply chain reliability, streamlining operations, reducing costs, and enhancing risk mitigation. AI-driven procurement automation has led to a 37% reduction in processing times and a 42% improvement in supplier selection accuracy, significantly outperforming traditional sourcing methods (IEEE, 2024). In Site Reliability Engineering (SRE) terms, AI is becoming a key enabler of resilient, scalable, and self-healing supply chain operations—critical for businesses operating in high-demand environments.

This session will explore how AI enhances procurement reliability through predictive analytics, autonomous decision-making, and risk mitigation strategies. Case studies from leading enterprises demonstrate 15-20% cost savings, with AI-driven spend analysis unlocking $3.8 million in annual optimizations (KPMG, 2024). AI-powered contract automation has resulted in a 60% reduction in processing time and a 30% decrease in compliance risks, while AI-driven early warning systems have achieved an 85% success rate in detecting supply chain disruptions—critical for SRE teams managing global infrastructure dependencies.

The discussion will also cover implementation challenges, such as data privacy (78% of organizations struggle with compliance) and AI integration hurdles with legacy systems (65% report difficulties) (IEEE, 2024). However, advances in AI-powered smart contracts and blockchain-backed procurement are projected to reduce procurement processing times by 85%, improving security, resilience, and compliance—aligning with SRE best practices.

Attendees will gain practical insights into AI-driven supply chain reliability, including sustainability initiatives that have cut corporate carbon footprints by 28% through AI-optimized logistics and supplier selection. This session will empower SRE professionals with strategies to leverage AI in procurement, ensuring cost efficiency, operational resilience, and reliability-driven supply chain management.

Parameswara Rao Tatini is a highly accomplished research scientist with over 20 years of industrial experience in designing, developing, and deploying machine learning (ML) and artificial intelligence (AI) applications at scale. His expertise spans Generative AI, deep and statistical machine learning, natural language processing (NLP), computer vision, and time-series forecasting. With a strong foundation in software engineering, signal processing, and data analytics, he has contributed significantly to AI-driven solutions across industries. His work has resulted in seven U.S. patents, highlighting his contributions to database management, AI-driven search, recommendation systems, and enterprise ML solutions. Currently, Parameswara serves as a Principal Data Scientist / Data Science Expert at SAP Ariba Inc., where he plays a crucial role in developing AI-powered enterprise solutions. He has led the end-to-end development of the Ariba Search application, leveraging Generative AI and Large Language Models (LLMs) to enhance search capabilities. His innovations include designing multi-lingual search classification models, improving search relevance by 8%, and developing an ML-powered recommendation system for Guided Buying, which earned a U.S. patent. Additionally, he pioneered the Pair Programming Assistant (PPA), an AI-driven tool that enhances coding efficiency through intelligent code suggestions, debugging support, and test case generation. His retail demand forecasting work reduced spoilage costs from 10% to 6%, saving $100M for a leading retailer. Before joining SAP Ariba, Parameswara spent a decade at IBM India Software Labs as an Advisory Software Engineer, where he spearheaded key architectural enhancements for IBM DB2. He was instrumental in developing automated database failover solutions, optimizing multi-tenant support for shared databases, and streamlining automated driver updates for client applications. His patented bi-directional data integration model significantly improved structured and unstructured data processing, making DB2 more efficient for enterprise applications. Parameswara holds a Master of Science in Software Systems and a Bachelor of Science in Information Systems from the Birla Institute of Technology and Science (BITS), Pilani, India. His technical skill set includes Python, SQL, R, MATLAB, Java, and cloud technologies like AWS, GCP, Azure, and SAP Business Technology Platform (BTP). Beyond his technical contributions, Parameswara is deeply invested in mentorship and leadership, guiding junior data scientists and collaborating with cross-functional teams to drive AI-powered innovation. His pioneering work in NLP, signal processing, and dimensionality reduction techniques (T-SNE, PCA, DCT, DWT) continues to push the boundaries of enterprise AI applications. Through his leadership and expertise, he remains at the forefront of AI, cloud computing, and data science innovations.

This session reveals the transformation journey of Microsoft's notification infrastructure that now powers Microsoft Teams, Copilot Chat, and Azure Communication Services. I'll share my experience leading the modernization of Azure Notification Hubs, which dramatically reduced costs while improving performance and reliability. The presentation examines how we migrated from legacy systems to a modern .NET core architecture while implementing essential resilience patterns like sophisticated retry mechanisms and circuit breakers to prevent cascading failures. I'll demonstrate how we maintained exceptional service uptime while processing billions of daily notifications across iOS, Android, Windows, and various OEM devices. You'll discover practical approaches to implementing latency-based routing, dynamic load balancing, and multi-region failover strategies that significantly improved performance under high traffic. I'll also discuss how we enhanced observability through Azure Monitor integration, balancing operational excellence with compliance requirements. This technical exploration provides architecture and engineering insights for those working with large-scale distributed systems, offering practical guidance for modernizing critical infrastructure while maintaining operational excellence. You'll gain actionable knowledge for building cloud-native notification systems that balance performance, cost, and reliability at enterprise scale.

Ankita Kamat is a Senior Software Engineer at Microsoft with over 10 years of experience in building scalable, high-performance systems. Currently leading development for Microsoft's Intelligent Conversation and Communications Cloud (IC3), she architects and implements messaging platforms that serve 300 million monthly active users across Microsoft Teams, Copilot Chat, and Azure Communication Services. Throughout her career at Microsoft, Ankita has demonstrated exceptional technical leadership and innovation. She played a key role in modernizing Azure Notification Hubs, achieving $3M in annual savings through strategic re-architecture. As a Tech Lead for the Account Notification Platform, she successfully led system design initiatives and cross-functional collaborations during a period of team transition. Prior to Microsoft, Ankita held senior engineering positions at Citigroup and UBS, where she specialized in developing enterprise-level applications using Angular, Spring MVC, and various front-end technologies. She has consistently shown expertise in optimizing performance, implementing secure solutions, and mentoring junior engineers. Ankita holds a Master of Science in Computer Science from Georgia Institute of Technology with a perfect 4.0 GPA, and a Bachelor's degree in Information Technology from Goa University. She is proficient in multiple programming languages including C#, F#, Java, and TypeScript, with extensive experience in cloud technologies like Microsoft Azure, Cosmos DB, and Kubernetes. A passionate advocate for operational excellence and innovation, Ankita has a proven track record of delivering impactful solutions that enhance user experiences while driving business growth. Her commitment to mentoring teams and streamlining development processes has consistently resulted in improved productivity and successful project outcomes.

What's the worst that can happen? Join me for a story about reliability, secrecy and potential 200k fatalities.