Cloud Dialogues

In this episode of Cloud Dialogues, Matt & Georgia dive into the big reveals and emerging trends from AWS re:Invent 2024. With over 72,000 attendees and an action-packed week in Las Vegas, this year's conference showcased a mix of cutting-edge announcements, practical innovations, and a few surprises. If you missed the event—or just want the TL;DR—this episode has you covered! What’s New at re:Invent? Conference Highlights - A record-breaking 72,000 registrations, with a strong in-person turnout in Las Vegas. - Expanded session formats, including breakout sessions, chalk talks, hands-on workshops, and a fresh addition: PeerTalk meetups for networking and knowledge sharing. Big AWS Announcements 1. Apple + AWS: A Powerful Partnership - Apple revealed it relies on AWS Compute for machine learning and AI applications (dubbed Apple Intelligence). - By leveraging AWS Graviton processors, Apple achieved a 40% efficiency boost. - Apple now operates across 10 AWS regions using Graviton-powered instances. 2. Smarter AI - Practical AI was the name of the game this year: - Bedrock Guardrails: Automated reasoning checks to reduce AI hallucinations. - Intelligent Prompt Routing: Dynamically directs prompts to the best AI model in the same family for the task. 3. Enterprise Enhancements - Declarative Policies: Simplified policy management across AWS environments. - Elastic VMware Service: Streamlines migrations from traditional VMware setups. - Amazon Q Developer: Packed with enterprise-friendly features like: - Automatic COBOL-to-modern-language code conversion. - Automated testing for smoother development. - Tools for investigating operational issues. 4. Data and Database Innovations - SageMaker Lakehouse: A unified hub for centralizing and managing data. - Aurora DSQL: A groundbreaking serverless, multi-region, Postgres-compatible database. Key Observations The hosts highlighted a noticeable shift from last year’s “AI everything” approach to more practical, value-driven AI solutions. The focus now is on delivering tools that make a tangible difference, and the hosts predict 2025 will bring realistic AI deployments as companies fine-tune their data and AI strategies. Looking Ahead The episode wraps with a teaser for the next recording in London, January 2025, and a call for listener input: who should be the next guest on Cloud Dialogues? Tell us in the comments!

In this exciting episode of Cloud Dialogues, Matt & Georgia sit down with Boris Tane—formerly of Baselime and now with Cloudflare — to dive deep into the latest in observability for cloud computing. Boris shares insights from his unique journey and discusses the challenges, requirements, and innovations shaping observability today. Boris combines technical expertise with a fresh perspective on observability. With a Master’s in Aerospace Dynamics and a background in predicting drone behavior, he’s tackled the complexities of data processing for unpredictable scenarios. His passion for analytics led him to a career in observability, where he now helps organizations unlock the potential of real-time data. What we Covered From Monitoring to Modern Observability - Traditional Monitoring: Built around dashboards that forecast issues or document past events. - Modern Observability: Emphasizes understanding application behavior live, without code changes, and goes beyond the classic "three pillars" (logs, metrics, traces) to focus on high-quality data and efficient query engines. Core Requirements for Effective Observability 1. High Cardinality: The ability to handle limitless unique values per field. 2. High Dimensionality: The capacity to track hundreds of attributes per event—like user ID, location, headers, and more. Tackling Common Obstacles - Over-reliance on Dashboards: Many teams are caught up in dashboards instead of implementing true observability. - Transition Hurdles: Moving from traditional logging to advanced observability can be tough. - Developer Experience: A seamless experience is crucial for adoption and successful implementation. Memorable Quotes “Observability without action is just storage. If you’re not actively using logs, metrics, and traces, you’re simply paying for storage.” “The quality of your observability is only as good as the quality of data your application produces.” Key Takeaways 1. Observability goes beyond debugging—it's about empowering teams to create better products. 2. Cloud providers are starting to integrate observability as a core platform feature. 3. Effective observability requires strong team ownership and a culture that values data-driven decisions. 4. A gradual, user-friendly transition to modern observability is essential—no need for complete application overhauls. Industry Trends - Cloud Maturity Levels: Different cloud providers vary widely in their observability capabilities. - Built-In Observability: Newer platforms like Cloudflare are incorporating high cardinality and high dimensionality into their infrastructure. - OpenTelemetry Standards: OpenTelemetry is emerging as a standard, but it still needs to be complemented with application-specific insights. Business Impact - Faster Problem Resolution: Enhanced observability enables teams to solve issues more efficiently. - Product Insights for Managers: Observability data can help track feature usage and user experience. - Informed Decision-Making: Teams with comprehensive data access make smarter choices. Final Thoughts This episode highlights that while the right tools are critical, the heart of effective observability lies in producing high-quality data and empowering teams to act on it. Organizations that foster a culture of observability and improve data quality over time will see a true business impact from their observability practices.

In this exciting episode of Cloud Dialogues, we are joined by Liz Fong-Jones, Field CTO at Honeycomb and former Google SRE, to explore the fascinating world of Site Reliability Engineering (SRE)—a game-changer for scaling and automating large systems. What We Covered: 1. Meet Liz Fong-Jones: Liz brings over a decade of SRE experience from her time at Google and Honeycomb, helping companies revolutionize how they manage reliability and automation. 2. The Origin Story: SRE actually predates the cloud! Born at Google in the early 2000s, SRE started as a way to automate manual system administration tasks and has since evolved into its own discipline, running parallel to DevOps. 3. SRE at Its Core:    - Minimize repetitive work (aka "toil") by automating everything you can.    - Use Service Level Objectives (SLOs) and Service Level Indicators (SLIs) to measure and maintain reliability. 4. Different SRE Models: There are different ways to implement SRE:    - Tools-based within platform teams    - Consultative SREs parachuting in to help teams    - Embedded SREs integrated within every team 5. The SRE Mindset: Curiosity and empathy are essential for SREs. Teams need a culture of psychological safety where concerns can be raised without fear. 6. The Magic of SLOs and SLIs: SLOs set reliability targets (like aiming for 99.5% uptime), while SLIs measure performance against those targets. Together, they ensure your systems are running smoothly. 7. FinOps Meets SRE: Liz explains how SREs can help balance reliability, performance, and costs using SLOs to allocate resources more efficiently. 8. Disaster Testing: Want proof SREs are ready for anything? Honeycomb regularly tests its disaster recovery by taking down an entire availability zone—on purpose! 9. Pro Tips for Executives: Thinking about implementing SRE at your company? Liz suggests starting with your biggest challenges, offering executive support, and setting clear, achievable SLOs. 10. Why Observability Matters: Observability is the backbone of SRE. Having real-time, actionable data is key for setting and managing effective SLOs. Plus, Liz gives covers off on her favorite ARM processors (for cost and environmental savings) and shares insights from her book Observability Engineering. This episode is a deep dive into SRE, filled with actionable insights and strategies for leaders looking to supercharge their reliability game. You won’t want to miss it!

In this episode of Cloud Dialogues, Matt & Georgia dive deep into the world of data strategy with Allison Howells, a seasoned expert with over 20 years of experience across industries like financial services, insurance, and consulting. Exploring the evolution of data platforms and the strategic decisions that help organizations harness the power of their data.  Here are some of the key topics the team breaks down: 1. Data Platforms: Then & Now – How data platforms have transformed over the past 10-15 years. 2. Aligning Data with Business – Why syncing your data strategy with your business goals is a game-changer. 3. Centralized vs. Decentralized Platforms – The pros and cons of each approach. 4. Data Governance & Quality – Why these are the cornerstones for driving successful AI and machine learning projects. 5. Data Context & Origin – The critical role of understanding where your data comes from and how it’s used. 6. Cultural Impact of Decentralized Platforms – How decentralization can foster innovation and shift organizational mindsets. Top Takeaways: - Data Strategy = Business Strategy: A successful data strategy is always grounded in solving real business problems. - Decentralized Data Platforms: These can enhance accountability, fuel innovation, and empower teams through data democratization. - Data Context is Key: Knowing the "where" and "why" behind your data ensures it’s used effectively. - Governance & Quality Matter: No matter your data platform, strong governance and data quality are essential. - Data Literacy is a Must: Every department needs to level up their data skills to thrive in today’s data-driven world. Curious about how to get your data initiatives off the ground and deliver business value from day one?  Allison shares actionable steps for executives to build a winning data strategy that drives results. To get the full executive playbook, tune in to the episode!

In this exciting episode of "Cloud Dialogues," Georgia and Matt return from their well-deserved breaks, ready to dive deep into the world of cloud operating models. They unpack the challenges organisations face when transitioning to cloud environments and share insights on how to navigate this complex landscape. Here's what you can expect: 1. Tackling Operating Model Challenges: Dive into the common pitfalls of cloud transformation, emphasizing how misaligned operating model changes often lead to trouble. 2. DevOps and Product-Centric Thinking: The shift from traditional IT operations to a DevOps and product-focused approach, stressing the need for a cultural shift within organisations. 3. Platform Engineering Essentials: The importance of customising cloud platforms to meet the unique needs of DevOps teams, all while creating an exceptional developer experience. 4. The Role of Cloud Centers of Excellence (COE): Discover how COEs bridge the gap between platform teams and developers, ensuring smooth and efficient cloud usage. 5. Evolution of Change Management: Learn how change management is transforming in cloud environments, moving from rigid approval processes to more flexible, enabling approaches. 6. FinOps and Accountability: The critical role of product owners in balancing cloud costs and value, making accountability a key focus. 7. Navigating Data Management: Challenges of managing data across organisations, proposing a federated model with a central data catalog. 8. AI and Platform Teams: Should organisations build their own AI capabilities or rely on existing SaaS solutions? The hosts offer their take on this hot topic. 9. Mastering Containerisation: They shed light on the complexities of managing containerised environments like Kubernetes, with practical insights. 10. Strengthening Security Operations: The importance of well-resourced security teams in understanding and mitigating risks within intricate cloud setups. 11. Transforming Operating Models: Finally, the hosts advocate for small, incremental changes over sweeping transformations, ensuring smoother transitions. To wrap things up, Georgia and Matt hint at an upcoming discussion on Site Reliability Engineering (SRE)—stay tuned!