Cloud Native Testing Podcast

In this episode of the Cloud Native Testing Podcast, host Ole Lensmar sits down with Alex Buccino, a 30-year tech industry veteran and former engineering leader at Google and Squarespace. The conversation dives deep into the monumental challenge of testing at Google scale, specifically focusing on the evolution of the Google Maps data infrastructure.

Alex introduces listeners to "SandMan" (Sandbox Manager), a revolutionary internal tool built on Google's Borg that allowed developers to spin up ephemeral, declarative test environments. By treating state and service components separately, SandMan enabled engineers to stand up mini-versions of massive, interconnected systems right in their own sandboxes. Whether you are currently using Kubernetes and Helm to manage your testing environments or navigating the complexities of microservice dependencies, this episode offers vital architectural lessons from one of the largest engineering organizations in the world.

Key Topics Covered:

• The Origin of SandMan: How the need to test the second implementation of Spanner led to a tool eventually adopted by tens of thousands of Google engineers.
• State vs. Service Components: The strategy of separating data and compute in test environments, allowing teams to seamlessly load sanitized production snapshots or empty databases via copy-on-write semantics.
• Defining "The Horizon": How to scope a test environment within a fully connected microservice graph using lightweight fakes, mocks, and null services.
• The Anti-Pattern of Startup Dependencies: Why forcing services to wait for databases to fully initialize before starting is a mistake, and how parallel provisioning dramatically speeds up sandbox creation.
• Declarative vs. Imperative Provisioning: Why relying on procedural scripts to build test environments becomes a massive bottleneck as you scale, and how declarative configurations solve the consistency problem across multiple instances.

In this episode of the Cloud Native Testing Podcast, host Ole Lensmar (CTO at Testkube) sits down with Lakshmi, one of the latest veterans to join the Testkube team. Together, they explore how the AI revolution is fundamentally changing the relationship between development and QA—and why testers might just be the ultimate "vibe coders."

Lakshmi shares his journey from electrical engineering and performance testing at Comcast to building his own AI-driven translation app without touching an IDE. The conversation dives deep into why the tester's mindset—always looking for edge cases, failure points, and hidden assumptions—makes them uniquely qualified to master natural language prompting and lead the charge in the age of AI.

In this episode, we discuss:

• The Evolution of Performance Engineering: How the shift to cloud auto-scaling mistakenly led teams to skip performance testing—and the massive cloud bills that brought it back.
• Why Testers Excel at Vibe Coding: How natural language has become the new programming interface, and why a QA mindset is the secret weapon for building resilient AI applications.
• Deterministic vs. Non-Deterministic Testing: Giving AI agents "personas" (like a 35-year-old mother looking for a mattress) to uncover unpredictable user paths in e-commerce.
• Moving Downstream with Failure Analysis: How Testkube is leveraging customizable agents to sift through logs, detect true flakiness, and move toward human-in-the-loop self-healing repositories.

Whether you're a developer curious about how AI is speeding up code deployment or a tester looking to claim your AI superpowers, this episode offers a fascinating look into the future of cloud-native quality assurance.

In this episode of the Cloud Native Testing Podcast, Ole Lensmar sits down with Sahana Nagabhushan to tackle the complexities of software lifecycle inefficiencies and developer productivity.

Drawing from her extensive product management experience at Fidelity Investments and Workday, Sahana breaks down why companies burn millions of dollars and years of engineering time building custom Internal Developer Portals (IDPs) like Backstage. Instead of relying solely on generic frameworks, she advocates for implementing a "production readiness layer"—a critical bridge between writing code and shipping it confidently to live environments.

Ole and Sahana also dive deep into the realities of AI in modern software development. They discuss why AI-generated code still desperately requires human oversight, the current limitations of AI in generating deterministic tests, and the strategic value of integrating existing top-tier tools (like Testkube) to maintain high-quality standards without reinventing the wheel.

Key Takeaways:

• The Cost of DIY Platforms: Why building internal developer portals from scratch can cost companies up to $16 million with delayed ROI.
• Defining Production Readiness: How companies can create a "trust layer" to ensure code is fully tested, governed, and truly ready for deployment.
• Backstage vs. Purpose-Built Tools: A candid discussion on why generic plugin frameworks might not solve core quality and testing issues.
• AI's Role in Testing: Why human context, sentiment, and oversight remain crucial when navigating AI-generated code and test creation.

Welcome to the first edition of the Cloud Native Testing Podcast for 2026! In this episode, host Ole Lensmar is joined by Yacine Kheddache to dive deep into Microcks, a CNCF Sandbox project dedicated to API mocking and simulation.

As cloud-native architectures grow more complex, the need to decouple services during development is critical. Yacine explains how Microcks serves as a "Swiss Army Knife" for developers, offering a single solution to mock and test REST, gRPC, GraphQL, and Event-Driven protocols (like Kafka and NATS). They discuss the tool's evolution from a centralized Kubernetes operator to a developer-friendly utility that runs natively in IDEs and pipelines, enabling true "shift left" testing.

Later in the conversation, they explore the intersection of API testing and Artificial Intelligence. Yacine details how Microcks is embracing the AI era by using Copilots to generate mock data and leveraging the Model Context Protocol (MCP) to make existing APIs accessible to LLMs.

Key Topics Discussed:

• The CNCF Journey: Microcks’ status as a community-driven Sandbox project.
• Polyglot Support: Mocking REST, GraphQL, gRPC, and AsyncAPI with one tool.
• The Testing Lifecycle: How to reuse mock data artifacts for automated contract and conformance testing in CI/CD.
• Shift Left: Moving testing from QA environments to local developer laptops and IDEs.
• AI & MCP: Generating datasets with AI and exposing APIs as tools for AI Agents using the Model Context Protocol.

In this episode, Ole Lensmar joins Richard Li, founder of Polar Sky, to explore how testing strategies must evolve for the age of AI. They discuss the journey of cloud-native tools like Telepresence and Richard's key insight for agentic coding: using automated tests as the "context" to guide AI behavior, rather than relying on complex prompts to describe system rules.

The conversation also digs into the mechanics of building AI-driven software, emphasizing why "evals" are critical for measuring success. Richard shares practical strategies for treating evaluations as data management problems and explains how his team uses simple AI agents to "babysit" CI pipelines and automatically resolve flaky tests.

Topics discussed:

• The Evolution of Telepresence: Moving from local debugging to headless execution in CI pipelines.
• Tests as Context: Why running a test suite is more effective than complex prompting for AI agents.
• Agentic Coding Strategy: Shifting focus from unit tests to integration and behavioral verification.
• AI Evals: Why evaluation is the most critical aspect of building reliable AI products.
• Babysitting CI: Using simple AI agents to identify and retry flaky tests automatically.

How do you shift left to ship fast when you work in a highly-regulated environment? In this episode, Thomas To explains how they built a modern testing culture to do just that. Drawing on his background at test automation company Katalon, Thomas details how "shift left" is primarily a cultural and mindset challenge, not just a technical one.

He shares practical insights on empowering developers to own quality, which in turn elevates the QA role to one of enablement and platform engineering. This cultural foundation becomes critical for shipping fast in a complex public safety domain, where testing in cloud-native environments involves navigating strict global regulations and data-access hurdles.

In this episode, you will learn about:

• What a "shift-left" culture really means: empowering developers and evolving the role of QA.
• How Axon's quality mindset is essential for navigating the complex testing challenges of a regulated, cloud-native environment.
• The practical use of AI and LLMs to facilitate the shift-left movement and help developers build quality in.

On this episode of the Cloud Native Testing podcast, Ole Lensmar welcomes Matt McLane from Doc Network to discuss how they completely transformed their development and testing lifecycle. Matt details their journey away from a slow, manual deployment process—which involved risky ad-hoc testing on a staging environment connected to production—to a modern, automated workflow built for quality and speed. By implementing tools like GitHub Actions, Argo CD, and Cargo, they built a pipeline where testing is a core component, not an afterthought.

The conversation dives deep into their cloud-native testing strategy. Matt explains how integrating Playwright for end-to-end tests created automated quality gates, giving them confidence with every deployment. He also shares their approach to non-functional testing with k6 to ensure their application can handle high-traffic events. A key highlight is their use of Tilt to spin up ephemeral environments for every pull request, enabling developers and QA to test features in isolation early in the process. This episode is a practical guide for anyone looking to build a robust testing culture supported by a modern CI/CD pipeline.

What's the connection between your performance tests and your monthly cloud bill? As it turns out, a lot. While many teams struggle with ever-growing Kubernetes costs, the solution isn't just about cutting expenses—it's about improving reliability.

In this episode, Ole is joined by Yasmine, COO of CloudBolt, to explore the complex world of Kubernetes rightsizing. Drawing from her experience at Stormforge, Yasmine unpacks the evolution from using performance testing to generate optimization data to the sophisticated machine learning models used today. This conversation goes beyond the tools to address the core challenges every organization faces when trying to tame its cloud-native environment.

Tune in to learn:

• The People Problem: Why there's natural friction between platform teams and developers on resource management, and how to bridge that gap.
• Reliability First: The crucial mindset shift of viewing rightsizing as a tool for reliability and performance, with cost savings as the byproduct.
• The Scale Challenge: Why manual scripts and "setting an average" are bound to fail when managing thousands of dynamic workloads.
• Testing for Confidence: How robust performance and functional testing can validate automated resource changes and give teams the confidence to deploy them.

In this episode, the main conversation is between host Ole Lensmar and guest Atul, who are joined by special guest Dmitry, the CEO of Testkube. Together, they dive deep into the evolving landscape of software testing, framed by Atul's fascinating journey from manual QA to a cloud native developer advocate.

The conversation kicks off with Atul sharing his unique career path, highlighting how starting in manual testing gave him a foundational understanding of system failure points—an invaluable edge in the world of development and cloud native technologies.

The core of the discussion explores the profound impact of AI on testing and development workflows. Atul points out a critical challenge: while AI can generate code at 10x the previous velocity, our CI/CD pipelines often become a bottleneck, as they weren't designed to handle such a high volume of pull requests. This leads to a crucial conversation about the need to not only scale testing efforts to ensure proper coverage for AI-generated code but also to scale the underlying CI/CD infrastructure itself. They also touch upon the significant software supply chain security concerns that arise when integrating AI agents into the development lifecycle.

Shifting focus to the broader ecosystem, the episode examines why testing often feels like an afterthought in the cloud native community, particularly at events like KubeCon. Atul argues that teams are often so focused on application functionality that they're "blinded" to the complexities of testing the entire infrastructure stack—from service meshes to container orchestration—until something breaks. The conversation also covers the nuanced practice of testing in production, balancing its benefits with the inherent risks.

In this illuminating episode, we pull back the curtain on Cloudflare's intricate approach to testing at a truly global scale. Join us as Sachin from Cloudflare's team shares invaluable insights into how they leverage their vast, distributed network to push the boundaries of reliability and performance. We'll explore the critical role of SLOs (Service Level Objectives) and SLIs (Service Level Indicators) in measuring Cloudflare's service reliability, and why error budgeting is essential for managing risk across their massive infrastructure.

Discover how Cloudflare embraces the often-debated, yet crucial, practice of testing in production to validate real-world performance. Learn how observability becomes their indispensable guide to understanding complex system behavior under immense load. We'll also discuss the power of automated monitoring in catching anomalies, the importance of collaborating with SLO teams within Cloudflare, and how continuous improvement is fueled by accurate metrics and data analysis gleaned from their unique testing environment. Whether you're a DevOps engineer, SRE, or simply curious about how an internet giant ensures its services stand strong, this episode offers actionable strategies and a fascinating look into Cloudflare's blueprint for resilience.