エピソード

  • How One Engineer Reduced a Monorepo CI Pipeline to 8 Minutes

    How One Engineer Reduced a Monorepo CI Pipeline to 8 Minutes
    2026/06/09
    Episode 40 of The Software Engineering Podcast dives into monorepo CI optimization. Lucas and Luna explore how a single engineer at a mid-sized tech company slashed continuous integration from 45 minutes to under 8 minutes by implementing incremental builds and dependency caching. They break down the specific techniques: Bazel-style action caching, remote caching with a shared Redis store, and a custom build graph that only rebuilds changed packages. The episode covers the before-and-after metrics, the hidden cost of full rebuilds, and how the team avoided common pitfalls like cache invalidation bugs and flaky test pollution. A concrete case study for anyone stuck with a slow monorepo pipeline. #Monorepo #CI #BuildOptimization #ContinuousIntegration #Bazel #IncrementalBuilds #DependencyCaching #Redis #BuildGraph #SoftwareEngineering #DevOps #Performance #CacheInvalidation #EngineeringCulture #TechPodcast #Technology #FexingoBusiness #BusinessPodcast Keep every episode free: buymeacoffee.com/fexingo
    続きを読む 一部表示
    9 分
  • How a Single Feature Flag Prevented a Five-Hour Outage

    How a Single Feature Flag Prevented a Five-Hour Outage
    2026/06/08
    In this episode, Lucas and Luna dive into the story of a fintech startup that avoided a catastrophic outage during a critical payment system migration — thanks to a single feature flag. They break down how the engineering team used a kill switch to gradually roll out a new database schema, caught a silent data corruption bug in the canary phase, and toggled back to the old system in under 90 seconds. The episode explores the architecture behind feature flags, the difference between a flag and a config, and why one senior engineer insisted on a global off-switch that saved the company from a five-hour incident. Lucas draws parallels to how companies like Netflix and Etsy use similar patterns for safe deployments. Listeners learn one concrete takeaway: every high-risk migration should start with a feature flag, not a flip. #FeatureFlags #SoftwareEngineering #IncidentResponse #Fintech #DeploymentStrategy #KillSwitch #DatabaseMigration #CanaryRelease #TechEngineering #SystemReliability #FexingoBusiness #BusinessPodcast #Technology #EngineeringBestPractices #ResilienceEngineering #SiteReliability #ProductionSafety #MigrationStrategy Keep every episode free: buymeacoffee.com/fexingo
    続きを読む 一部表示
    14 分
  • How One Engineer Automated Code Review with a Custom Static Analysis Tool

    How One Engineer Automated Code Review with a Custom Static Analysis Tool
    2026/06/08
    In this episode, Lucas and Luna explore how a single engineer at a mid-size SaaS company built a custom static analysis tool that automated 80% of code review comments. They break down the specific rules the tool enforced, how it was integrated into the CI pipeline, and the surprising cultural pushback the engineer faced. The episode includes a real-world example: a rule that caught unnecessary database queries in pull requests, saving hours of manual review per week. Lucas and Luna also discuss the trade-offs between custom tools and off-the-shelf linters like ESLint and SonarQube. Perfect for senior engineers and tech leads looking to reduce review bottlenecks without sacrificing code quality. #StaticAnalysis #CodeReview #Automation #EngineeringProductivity #CustomTools #CI_Pipeline #Linter #DeveloperExperience #TechLead #SaaS #PullRequest #DatabaseQueries #SoftwareEngineering #Technology #FexingoBusiness #BusinessPodcast #CodeQuality #TechPodcast Keep every episode free: buymeacoffee.com/fexingo
    続きを読む 一部表示
    7 分
  • How One Engineer Cut Incident Response from Hours to Seconds with a Runbook

    How One Engineer Cut Incident Response from Hours to Seconds with a Runbook
    2026/06/07
    Episode 37 of The Software Engineering Podcast with Fexingo dives into a specific operational win: how a senior engineer at a mid-size fintech company automated incident response runbooks, slashing mean time to resolution from over two hours to under thirty seconds. Lucas and Luna walk through the before-and-after — the chaotic Slack threads, the manual playbook that lived in a Google Doc, and the gradual shift to code-driven remediation. They discuss why a runbook-as-code approach reduced human error, how the team tested incident flows in staging, and the one misstep that nearly caused a false positive cascade. The episode also touches on the broader movement toward 'incident response as software' and what it means for on-call culture. No hot takes, no buzzwords — just a concrete story of making systems more resilient by writing better automation scripts. #IncidentResponse #RunbookAutomation #SiteReliabilityEngineering #DevOps #OnCall #IncidentManagement #SoftwareEngineering #Automation #RunbookAsCode #Fintech #EngineeringCulture #MTTR #ReliabilityEngineering #Observability #Postmortem #Technology #FexingoBusiness #BusinessPodcast Keep every episode free: buymeacoffee.com/fexingo
    続きを読む 一部表示
    7 分
  • How a Single Bit Flip Brought Down an Entire Data Center

    How a Single Bit Flip Brought Down an Entire Data Center
    2026/06/07
    In episode 36 of The Software Engineering Podcast with Fexingo, Lucas and Luna dive into one of the most infamous hardware-induced software bugs in recent memory: the 2021 Facebook outage caused by a single bit flip. Lucas explains how a routine configuration change triggered a cascading failure that took down Facebook, Instagram, and WhatsApp for over six hours. He walks through the exact sequence — a BGP withdrawal, DNS failures, and a data center network meltdown — and why a single incorrect bit in a router's memory was the root cause. Luna challenges the conventional wisdom about redundancy and asks whether engineers can realistically guard against single-bit errors at scale. They discuss cosmic rays, memory error-correcting codes, and the trade-offs between software abstraction and hardware reality. Along the way, they share practical lessons for engineers designing resilient systems: from careful change management to the dangers of assuming hardware is perfect. #SoftwareEngineering #Technology #FacebookOutage #BitFlip #BGP #DNS #DataCenter #Resilience #Infrastructure #Networking #CosmicRays #ECC #ErrorCorrection #ChangeManagement #CascadingFailure #EngineeringLessons #FexingoBusiness #BusinessPodcast Keep every episode free: buymeacoffee.com/fexingo
    続きを読む 一部表示
    6 分
  • How One Engineer Slashed Build Times from 40 Minutes to 90 Seconds

    How One Engineer Slashed Build Times from 40 Minutes to 90 Seconds
    2026/06/06
    In this episode of The Software Engineering Podcast, Lucas and Luna dive into the story of a senior engineer at a mid-size SaaS company who cut their CI build pipeline from 40 minutes to 90 seconds. They walk through the specific bottlenecks: a monolithic Gradle build with unnecessary task dependencies, Docker layers being rebuilt on every commit, and a test suite that ran sequentially. The fix was a combination of incremental compilation, parallel test execution, and a custom caching layer using BuildKit. They also discuss the cultural resistance to changing a 'working' build system and how the engineer used data to win buy-in. By the end, listeners get a concrete playbook for auditing their own CI pipelines and a reminder that build speed is often the cheapest performance optimization you can make. #SoftwareEngineering #BuildOptimization #CI #Gradle #Docker #BuildKit #IncrementalCompilation #ParallelTesting #DevEx #DeveloperProductivity #CacheStrategy #BuildPipeline #Tech #EngineeringBestPractices #FexingoBusiness #BusinessPodcast #TheSoftwareEngineeringPodcast #Fexingo Keep every episode free: buymeacoffee.com/fexingo
    続きを読む 一部表示
    8 分
  • How One Engineer Cut Docker Image Size by 90 Percent

    How One Engineer Cut Docker Image Size by 90 Percent
    2026/06/06
    Episode 34 of The Software Engineering Podcast with Fexingo dives into a practical optimization story: how one engineer at a mid-sized SaaS company shrank a bloated Docker image from 2.1 GB to just 210 MB. Lucas and Luna walk through the specific techniques used — from switching to Alpine base images to eliminating layer bloat with multi-stage builds and removing unnecessary packages. They discuss why smaller images matter beyond disk space: faster CI pipelines, reduced network transfer, and improved security posture. The episode also touches on the engineer's debugging process, including how they used 'docker history' and 'dive' to find hidden culprits like cached pip packages and leftover build artifacts. Listeners will walk away with a concrete checklist they can apply to their own Dockerfiles today. #Docker #ContainerOptimization #AlpineLinux #MultiStageBuild #DevOps #CI #SaaS #SoftwareEngineering #Technology #FexingoBusiness #BusinessPodcast #EngineeringBestPractices #Dockerfile #LayerCaching #ImageSize #Performance #Security #Tooling Keep every episode free: buymeacoffee.com/fexingo
    続きを読む 一部表示
    8 分
  • How One Engineer Refactored a 10 Year Old Codebase in Six Weeks

    How One Engineer Refactored a 10 Year Old Codebase in Six Weeks
    2026/06/05
    Episode 33 of The Software Engineering Podcast with Fexingo. Lucas and Luna dive into the story of a senior engineer who inherited a monolithic ten-year-old codebase with zero tests and a single deployment causing multi-hour outages. Over six weeks, they systematically added integration tests, extracted domain modules, and built a CI pipeline that cut deployment failures by 90 percent. The episode focuses on the practical first steps: the decision to start with a single endpoint, the use of characterization tests to capture existing behavior, and the trade-off between perfect abstraction and incremental improvement. No hot takes – just the concrete tactics that turned a legacy nightmare into a deployable system. #LegacyCodebase #Refactoring #IntegrationTesting #CI #CharacterizationTests #IncrementalImprovement #SoftwareEngineering #CodeArchitecture #BestPractices #TechnicalDebt #DeploymentPipeline #TestingStrategy #ModuleExtraction #SeniorEngineer #FexingoBusiness #BusinessPodcast #Technology #Episode33 Keep every episode free: buymeacoffee.com/fexingo
    続きを読む 一部表示
    9 分