constakinola={currentFocus: "Screamwomen, VeloDB & ShipDeck",learning: ["Distributed Systems","Database Internals","OS Kernel & Memory Management"],looking: "Open Source Collaboration on Systems Infrastructure 🤝",stack: {frontend: ["TypeScript","React.js","Next.js","React Native"],backend: ["Node.js","Java (JDK 17/21)","Express","NestJS","PostgreSQL","Prisma"],systems: ["Off-Heap Memory Management","SIMD (Vector API)","Disruptor Ring Buffers","Hashed Wheel Timers"],cloud: ["Docker","AWS","Nginx","GitHub Actions","CI/CD"]},contact: "ayobamiakinola84@gmail.com",funFact: "I believe the best Garbage Collector is no heap allocations at all! 💾"};- 🔭 I'm currently working on Screamwomen (secure multi-tenant platform) and custom systems engines (VeloDB, ShipDeck, Hydra, Chronos).
- 🌱 I'm currently exploring Distributed Systems Consensus and Low-Latency Database Engineering.
- 👯 I'm looking to collaborate on Open-source Systems & Infrastructure projects.
- 💬 Ask me about NextJS, TypeScript, Java Systems Programming, Off-heap Memory, Concurrency, and Container Proxies.
- 📫 How to reach me ayobamiakinola84@gmail.com
- 🚀 ShipDeck: A self-hosted Developer PaaS (like Heroku) written in TypeScript and Java. Automates Git-to-deploy pipelines, handles dynamic TCP reverse proxying, atomic blue-green container swaps via Docker socket interactions, and collects cgroups system telemetry.
- 🔥 VeloDB: An in-memory vector similarity database written in Java and TypeScript. Achieves sub-10ms search query speeds over 500K+ vectors by bypassing JVM Garbage Collection pauses entirely using
sun.misc.Unsafeoff-heap memory, accelerated by hardware-level SIMD math via the Java Vector API. - 🌊 Hydra: A high-throughput log aggregator routing 50K+ requests/sec. Features edge token-bucket rate limiters in TypeScript, passing log streams to a lock-free circular Ring Buffer in Java to execute zero-allocation asynchronous batch writes.
- ⏰ Chronos: A fault-tolerant distributed cron task scheduler. Uses Hashed Wheel Timers in Java to schedule millions of tasks in
$O(1)$ constant time, dispatching execution workloads to TypeScript worker nodes over custom keep-alive TCP socket channels.
