Episode 85 of The Technical Co-Founder Podcast explores how a seed-stage startup called EdgeBase replaced their entire client-side compute layer by compiling Go directly to WebAssembly. Lucas and Luna unpack why they chose Go over JavaScript for a browser-based WebAssembly module, how they handled memory management and cross-language function calls, and the surprising performance results — near-native speeds for data processing tasks. They also discuss the trade-offs: debugging difficulties, bundle size bloat, and when WebAssembly actually makes sense versus a traditional backend API. Plus, a candid moment about how listener support via buy me a coffee dot com slash fexingo keeps the show ad-free and focused on real engineering stories.

#WebAssembly #GoLanguage #Wasm #BrowserCompute #EdgeBase #StartupEngineering #CTOLife #TechnicalCoFounder #BusinessAndTechnology #SoftwareArchitecture #Performance #Debugging #Compilation #MemoryManagement #GoToWasm #Frontend #Backend #FexingoBusiness

Keep every episode free: buymeacoffee.com/fexingo

Lucas and Luna dive into how one startup uses semantic versioning (SemVer) to manage dependencies across 150 microservices without breaking production. They walk through the specific strategy: MAJOR.MINOR.PATCH tags on every npm package, automated pre-release validation, and a bot that blocks incompatible upgrades. The episode centers on the real numbers: before SemVer, the team faced 12 production incidents per quarter from version mismatches; after adopting a strict SemVer pipeline with lockfiles and canary deploys, that dropped to zero incidents for seven consecutive quarters. Lucas explains the trade-offs: the overhead of bumping versions versus the cost of unexpected breakage. Luna pushes back on whether SemVer is still relevant in a world of monorepos and trunk-based development. The discussion lands on a pragmatic middle ground — use SemVer for external-facing packages, and internal coherency for private modules.

#SemanticVersioning #SemVer #DependencyManagement #Microservices #StartupEngineering #DevOps #PackageManagement #Versioning #ContinuousIntegration #Nodejs #Npm #BreakingChanges #Monorepo #SoftwareEngineering #TechStartup #BusinessAndTechnology #FexingoBusiness #BusinessPodcast

Keep every episode free: buymeacoffee.com/fexingo

In this episode, Lucas and Luna explore how a fintech startup called PayShield implemented rate limiting to prevent API abuse after a credential-stuffing attack compromised 2,000 accounts in one weekend. They break down the technical architecture — token bucket vs. sliding window, per-user vs. per-endpoint limits — and the business tradeoff: too aggressive and you block legitimate users, too loose and you invite fraud. PayShield's CTO shares how they reduced abuse by 95 percent while keeping false positives under 0.1 percent. Lucas and Luna discuss how this approach scales from a single Redis instance to a distributed cluster, and why rate limiting is a first-class feature, not an afterthought.

#RateLimiting #APIAbuse #PayShield #TokenBucket #SlidingWindow #Redis #Fintech #Security #Engineering #Startup #TechnicalCoFounder #Fexingo #BusinessPodcast #SoftwareEngineering #AbusePrevention #DevOps #API #Scalability

Keep every episode free: buymeacoffee.com/fexingo

Lucas and Luna dive into event sourcing, a data architecture pattern that stores every state change as an immutable event. They explore how one B2B SaaS startup used event sourcing to build a full audit trail, recover from data corruption in minutes, and enable time-travel debugging for customer support. The episode walks through the trade-offs: why the team chose PostgreSQL as their event store, how they handled schema evolution over 18 months, and why they decided against Kafka for this use case. Listeners will learn what event sourcing feels like in practice, when it's worth the complexity, and the one question every team should ask before adopting it.

#EventSourcing #DataArchitecture #PostgreSQL #CQRS #StartupEngineering #AuditTrail #TimeTravelDebugging #ImmutableData #SchemaEvolution #B2BSaaS #EventStore #SoftwareArchitecture #TechPodcast #Fexingo #BusinessAndTechnology #EngineeringFounders #CTOInsights #CodeToProduct

Keep every episode free: buymeacoffee.com/fexingo

In Episode 81 of The Technical Co-Founder Podcast, Lucas and Luna explore how a fast-growing fintech startup uses chaos engineering to test its production systems without breaking customer trust. They break down the specific tools the team deploys — including LitmusChaos and Gremlin — and explain how the startup runs weekly 'chaos hours' where it deliberately injects latency, kills pods, and simulates region failures. Lucas shares how the team measures resilience using Service Level Objectives and a custom 'blast radius' dashboard, and why the CTO decided to start with read-only endpoints before touching the payment flow. Luna challenges whether chaos engineering is just 'breaking things for fun' or a genuine risk-reduction strategy. The episode also covers how the startup convinced its CEO to approve the initiative by tying chaos experiments to a 30 percent reduction in PagerDuty alerts and a 50 percent drop in mean time to recovery. No hype, just a real-world case study in building confidence through controlled failure.

#ChaosEngineering #ResilienceTesting #ProductionTesting #LitmusChaos #Gremlin #SiteReliabilityEngineering #SRE #ServiceLevelObjectives #SLOs #Fintech #StartupEngineering #FailureTesting #BlastRadius #PagerDuty #MeanTimeToRecovery #BusinessAndTechnology #FexingoBusiness #BusinessPodcast

Keep every episode free: buymeacoffee.com/fexingo

Episode 80 of The Technical Co-Founder Podcast — Lucas and Luna explore how a 25-person B2B SaaS startup called LayerVault replaced traditional security tooling with developer-led practices. They walk through LayerVault's adoption of Sigstore for software supply chain signing, OpenSSF Scorecards for open-source dependency hygiene, and a lightweight threat modeling ritual that engineers run as part of their sprint planning. Lucas breaks down how the company reduced its mean time to remediate critical vulnerabilities from 14 days to under 48 hours without hiring a single dedicated security hire. Luna challenges whether this approach scales past 50 engineers and presses on the cultural prerequisites. The episode closes with a reflection on when developer-led security becomes a bottleneck rather than an enabler.

#LayerVault #Sigstore #OpenSSF #DeveloperLedSecurity #ThreatModeling #SoftwareSupplyChain #ShiftLeft #B2BSaaS #EngineeringCulture #VulnerabilityManagement #SprintPlanning #StartupSecurity #BusinessAndTechnology #TechnicalCoFounder #FexingoBusiness #BusinessPodcast #SecureDevelopment #DevSecOps

Keep every episode free: buymeacoffee.com/fexingo

In this episode, Lucas and Luna explore how a fast-growing B2B SaaS company called DataSync avoided a catastrophic production outage by catching a silent schema drift before it reached production. They walk through the specific tooling — a combination of declarative migrations, CI checks, and a custom drift detector that compares the intended schema against the live database on every deploy. The case study shows how the team reduced deployment-related database incidents from five per quarter to zero over six months. Lucas explains why traditional migration scripts are insufficient when multiple engineers ship database changes in parallel, and how treating schema as code with automated verification closes a critical gap in the deployment pipeline. The episode also touches on the broader principle: as organizations adopt trunk-based development and deploy many times a day, database schema management becomes the bottleneck — and drift detection is the overlooked solution.

#SchemaDrift #Database #DataSync #B2BSaaS #DevOps #CI #TrunkBasedDevelopment #Migration #TechOps #DatabaseReliability #Infrastructure #SoftwareEngineering #Startup #BusinessAndTechnology #FexingoBusiness #BusinessPodcast #Engineering #PostgreSQL

Keep every episode free: buymeacoffee.com/fexingo

Episode 78 dives into a specific technical decision: how one early-stage infrastructure startup replaced its sidecar proxies with WebAssembly modules running in the data plane. Lucas and Luna walk through the company's original architecture—a standard Envoy sidecar per pod—and the problems that emerged as they scaled to 200 microservices: memory overhead, version drift, and cold-start latency. They explain the shift to a Wasm-based filter chain, how it cut memory-per-node by 60 percent and reduced deployment coupling between platform and product teams. Luna raises the operational trade-off: Wasm debugging is still immature. Lucas points to the emerging Wasm ecosystem—specifically the bytecode alliance's runtime and the rise of Wasm-native API gateways. The episode ends with a reflection on when a sidecar replacement makes sense and when it doesn't, using the startup's specific metrics as the guide. If you're evaluating service mesh alternatives or wondering where WebAssembly fits in production infrastructure, this episode gives you one grounded example to start from.

#WebAssembly #SidecarProxy #Envoy #ServiceMesh #CloudNative #Infrastructure #Startup #Microservices #PlatformEngineering #BytecodeAlliance #WasmRuntime #DataPlane #PodArchitecture #ColdStart #MemoryOptimization #BusinessAndTechnology #FexingoBusiness #BusinessPodcast

Keep every episode free: buymeacoffee.com/fexingo

In this episode of The Technical Co-Founder Podcast, Lucas and Luna dive into eBPF (extended Berkeley Packet Filter), a technology that lets developers run sandboxed programs in the Linux kernel without changing kernel code or loading modules. They explore how a startup called Pixie (acquired by New Relic) built a no-instrumentation observability platform using eBPF, enabling engineers to trace every system call, network packet, and function call with zero code changes. Lucas breaks down how eBPF works at a high level—attaching hooks to kprobes and tracepoints, running in a verified virtual machine—and why it's a game-changer for performance monitoring, security, and networking. Luna asks about the learning curve and trade-offs, including kernel compatibility and safety concerns. The episode includes a brief, natural donation segment tied to building better tools. If you're a developer curious about kernel-level insights without the overhead, this episode is for you.

#eBPF #Kernel #Observability #Pixie #NewRelic #Linux #PerformanceMonitoring #SysAdmin #DevOps #SRE #CloudNative #Startup #Engineering #Technology #Podcast #FexingoBusiness #BusinessPodcast #TheTechnicalCoFounder

Keep every episode free: buymeacoffee.com/fexingo

Episode 76 of The Technical Co-Founder Podcast digs into how a mid-stage B2B SaaS company called SyncLayer used GraphQL Federation to unify 20 independent microservices into a single data graph — without rewriting any backend. Lucas walks through the specific problem: frontend teams were making 8 to 12 round trips per page load, and the API gateway had become a bottleneck with spaghetti logic. Luna asks about the migration strategy — did they go big-bang or incremental? Lucas explains how they used Apollo Federation with a 'strangler fig' approach, migrating one subgraph at a time over six months. The result: 60 percent fewer network calls, faster feature development, and one unified schema documentation that made onboarding 70 percent faster for new engineers. The episode also covers the unexpected win: the federated graph gave them built-in tracing because the Apollo router automatically propagates OpenTelemetry headers across subgraphs. No magic, just solid engineering trade-offs.

#GraphQL #Federation #Apollo #Microservices #API #B2B #SaaS #TechStack #DeveloperExperience #SoftwareEngineering #BusinessAndTechnology #FexingoBusiness #BusinessPodcast #TechnicalCoFounder #EngineeringLeadership #Backend #Frontend #Migration

Keep every episode free: buymeacoffee.com/fexingo