Lucas and Luna dive into the technical architecture behind Figma's real-time collaboration engine. They explore how Figma moved from a standard web app to a custom multi-user sync layer using Operational Transform and CRDT-inspired techniques, handling hundreds of simultaneous designers editing the same vector without conflict. With specific numbers on latency and data structures, they break down the core challenge: making a complex design tool feel as snappy as a local app while running entirely in the browser. Luna pushes on edge cases like offline editing and conflict resolution, and Lucas explains Figma's migration from WebGL 1 to WebGL 2 for rendering performance. The episode also covers how Figma handled the 'onboarding cliff' when enterprise teams brought 50+ designers into a single file. Listeners will learn one concrete thing: why Figma chose to build its own implementation instead of using existing CRDT libraries, and the trade-offs involved.

#Figma #RealTimeCollaboration #CRDT #OperationalTransform #WebGL #BrowserEngine #DesignTools #Multiplayer #VectorGraphics #ConflictResolution #Latency #Architecture #Engineering #BusinessAndTechnology #TechnicalLeadership #FexingoBusiness #BusinessPodcast #CTOPodcast

Keep every episode free: buymeacoffee.com/fexingo

In this episode, Lucas and Luna dive into how Stripe rebuilt its core payments engine to handle compliance, currency, and latency across over 100 countries. They explore the architectural decisions behind Stripe's global payment infrastructure, including the unified API that abstracts local banking regulations, the edge-caching layer for 4x faster checkout in emerging markets, and the team structure that allows Stripe to ship country-specific features weekly without breaking existing integrations. The hosts also discuss the trade-offs between centralized routing logic and localized processing, and how Stripe's engineering org evolved to support this scale. A concrete look at the hidden complexity behind one of the internet's most critical platforms.

#Stripe #PaymentInfrastructure #GlobalPayments #API #Fintech #Scale #Engineering #Architecture #CTO #TechnicalLeadership #BusinessAndTechnology #FexingoBusiness #BusinessPodcast #StripePayments #Compliance #EdgeCaching #UnifiedAPI #EngineeringOrg

Keep every episode free: buymeacoffee.com/fexingo

In this episode of The CTO Podcast, Lucas and Luna dive into how Loom rebuilt its video recording and streaming engine to handle over 200 million videos per year while keeping latency under one second for uploads and playback. They explore the architectural shift from monolithic encoding to a microservices pipeline using WebAssembly for client-side transcoding, the trade-offs of moving from FFmpeg to a custom encoder, and how Loom cut infrastructure costs by 40 percent while improving reliability. The hosts also discuss the challenges of supporting asynchronous communication at scale, including real-time trimming and seamless playback across devices. If you're building a video platform or optimizing media delivery, this episode offers concrete lessons on balancing performance, cost, and user experience. Plus, Lucas shares how the team used Rust to reduce CPU usage for encoding by 60 percent. Tune in for a deep technical breakdown of one of the most popular async video tools in the workplace.

#Loom #AsyncCommunication #VideoStreaming #WebAssembly #Rust #Transcoding #Microservices #FFmpeg #CostOptimization #Latency #Engineering #CTO #TechLeadership #Architecture #BusinessAndTechnology #FexingoBusiness #BusinessPodcast #MediaDelivery

Keep every episode free: buymeacoffee.com/fexingo

Lucas and Luna explore how Discord moved from a standard WebRTC stack to a custom UDP transport layer, achieving sub-50ms latency and near-zero packet loss during peak gaming events like Call of Duty launches. They break down the engineering decisions behind Discord's transition from a hobbyist chat app to the backbone of 200 million monthly active users, focusing on the trade-offs between reliability and speed. The episode dives into how Discord's team handled jitter buffers, forward error correction, and the shift from Google's libjingle to their own 'Krisp' noise suppression. A concrete look at real-time communication at scale.

#Discord #VoiceEngine #WebRTC #Sub50msLatency #ZeroPacketLoss #UDPMux #KrispNoiseSuppression #CallOfDutyLaunch #200MillionUsers #RealtimeCommunication #Scalability #EngineeringCulture #BusinessAndTechnology #FexingoBusiness #BusinessPodcast #CTOPodcast #TechnicalLeadership #Architecture

Keep every episode free: buymeacoffee.com/fexingo

Lucas and Luna dive into the technical decisions behind Airbnb's 2024 search architecture overhaul. With 8 million active listings and billions of queries daily, Airbnb moved from a monolithic Elasticsearch cluster to a multi-tiered retrieval pipeline combining inverted indexes, vector embeddings, and real-time ranking. They explore why Airbnb chose to build custom microservices over managed solutions, how they reduced median search latency by 40 percent without sacrificing relevance, and the operational headaches of migrating 500-plus engineers to a new query language. A concrete look at search infrastructure at planetary scale.

#Airbnb #SearchArchitecture #Elasticsearch #VectorSearch #Microservices #LatencyOptimization #EngineeringLeadership #TechInfrastructure #BusinessAndTechnology #CTO #SystemDesign #Scalability #MachineLearning #RealTimeRanking #DataEngineering #MigrationStrategy #FexingoBusiness #BusinessPodcast

Keep every episode free: buymeacoffee.com/fexingo

In this episode, Lucas and Luna dive into the specific technical decisions GitLab made to scale its CI/CD platform from tens of thousands of pipelines to over one million per day. They focus on the choice of a monolithic application architecture, the use of PostgreSQL partitioning and Redis Cluster, and the trade-offs around job scheduling and queue management. The discussion highlights how GitLab avoided microservice complexity early on and instead optimized a single codebase to handle massive parallel workloads. Listeners will learn about the concept of 'fleet-wide coordination' and how GitLab's engineering team used feature flags and gradual rollout to maintain reliability during rapid growth. This is a concrete case study in scaling developer tools without rewriting everything.

#GitLab #CICD #PipelineScaling #MonolithArchitecture #PostgreSQL #RedisCluster #JobScheduling #FeatureFlags #DeveloperTools #ContinuousIntegration #ContinuousDeployment #EngineeringCulture #TechnicalLeadership #BusinessAndTechnology #DevOps #Scale #FexingoBusiness #BusinessPodcast

Keep every episode free: buymeacoffee.com/fexingo

Slack's search is used over 10 million times a day. But the original Elasticsearch-based index couldn't keep up as the platform grew to billions of messages. In this episode, the CTO Podcast breaks down how Slack's engineering team rebuilt its search infrastructure from the ground up — migrating from a single Elasticsearch cluster to a custom search engine called 'Search It'. We cover why they moved away from Elasticsearch, how they sharded data across hundreds of nodes, and the key trade-offs they made to keep search latency under 200 milliseconds. Lucas and Luna also discuss the human side of a multi-year migration: getting buy-in from leadership, managing developer burnout, and the moment the new search went fully live. If you're responsible for scaling a product that depends on fast, reliable search, this episode is packed with practical lessons.

#Slack #SearchEngine #Elasticsearch #Scalability #Infrastructure #Migration #Engineering #CTO #TechLeadership #Architecture #Productivity #Business #Technology #FexingoBusiness #BusinessPodcast #CTOPodcast #RealTimeSearch #Backend

Keep every episode free: buymeacoffee.com/fexingo

Lucas and Luna dive into Notion's technical journey to rebuild its block-based sync engine for offline-first and real-time collaboration. They explore the architectural shift from optimistic local writes to a hybrid local-sync model, the challenges of conflict resolution with operational transforms, and how Notion handled state reconciliation for millions of users. Specific numbers: over 100 million blocks synced daily, sub-second latency for local writes, and a 40% reduction in sync errors post-migration. The episode also touches on the trade-offs between local-first and server-authoritative models, and what other builders can learn from Notion's engineering decisions.

#Notion #OfflineFirst #SyncEngine #LocalFirst #OperationalTransform #ConflictResolution #RealTimeCollaboration #BlockEngine #Engineering #Architecture #TechLeadership #CTO #Business #Technology #FexingoBusiness #BusinessPodcast #Scaling #SoftwareEngineering

Keep every episode free: buymeacoffee.com/fexingo

Notion's block-level architecture was designed for single-user documents. But as the product scaled to 100 million users, real-time collaboration exposed deep cracks in its sync model. This episode unpacks how Notion's engineering team rebuilt its block engine from the ground up — migrating from a naive operational transform approach to a custom hybrid model that blends local-first performance with server-authoritative sync. We walk through the specific tradeoffs they made: why they ditched CRDTs for a timestamped delta approach, how they handled conflict resolution for nested blocks without locking, and the surprising lesson about battery life that shaped their sync cadence. If you're building a collaborative product or thinking about offline-first architecture, this episode has a concrete case you can reference.

#Notion #BlockEngine #Collaboration #LocalFirst #SyncArchitecture #OperationalTransform #CRDT #ConflictResolution #OfflineFirst #EngineeringOrg #TechnicalLeadership #CTO #ProductivitySoftware #RealTimeSync #Architecture #FexingoBusiness #BusinessPodcast #BusinessAndTechnology

Keep every episode free: buymeacoffee.com/fexingo

Episode 69 of The CTO Podcast. Lucas and Luna dive into how MongoDB's engineering team redesigned their core storage engine to handle multi-terabyte workloads without sacrificing developer velocity. They break down the specific challenge: a single customer complaint about a 12-terabyte shard that took 47 hours to rebalance. That complaint triggered a two-year effort to rewrite the WiredTiger storage engine's write-ahead log and compression layer. Lucas explains why the old architecture hit a wall at around 8 terabytes per node, and how the team's decision to switch from per-document compression to page-level compression made rebalancing 40x faster. Luna brings in a parallel from her own experience at a fintech startup that hit similar scaling pains. They discuss the trade-offs the MongoDB team made—accepting a 15 percent increase in storage cost per gigabyte in exchange for predictable rebalance times. A concrete look at how one of the most popular NoSQL databases evolved to meet the demands of modern workloads.

#MongoDB #WiredTiger #NoSQL #StorageEngine #DatabaseArchitecture #CTOPodcast #TechnicalLeadership #EngineeringOrgs #Scaling #Compression #WriteAheadLog #Rebalancing #MultiTerabyte #FexingoBusiness #BusinessPodcast #BackendEngineering #DatabaseInternals #ProductionEngineering

Keep every episode free: buymeacoffee.com/fexingo