ERAYACK.

About

Infrastructure, Distributed Systems, and Causal Consistency.

I build production platforms where reliability, throughput, and distributed state alignment are treated as first-class design requirements from day one.

Deployment Cycle

Hours → Minutes

Focus

High-Precision Distributed State

Current Direction

Kernel-Integrated Admission & Causal Order

Hi, I'm Eray. Senior Infrastructure / Platform Engineer building cloud-native systems with a focus on high-fidelity reliability. I operate at the intersection of kernel-level performance (eBPF) and edge-level consistency, ensuring that high-stakes applications, specifically in ML and Energy, remain deterministic under load.

Recently, I developed a Causal Consistency Toolkit in Rust and Lua to solve the "stale-read" problem in distributed inference. This includes Chronos, a high-frequency sequencer for manufacturing causal order, and Vortex, a decentralized membership engine. My approach is rooted in Simulation-First Engineering; I do not just deploy systems, I build deterministic chaos harnesses to prove their failure boundaries before they hit production.

Previously, at Semiotic Labs / The Graph, I worked on performance-critical systems (p99 latency, caching, throughput) and contributed to the major The Graph 2.0 Horizon upgrade, operating at the leading edge of distributed indexing protocols.

I use Rust where determinism and performance are non-negotiable infrastructure constraints. My toolkit is biased toward zero-unsafe systems and strict error handling, reflecting an engineering philosophy that values system safety as much as system speed.

Operating Model

Build for Causal Order

Design around monotonic state, deterministic sequencing, and fault-tolerant membership.

Quantify the Jitter

Use deterministic simulation to move from "best-effort" to "mathematically-verifiable" reliability.

Treat Infrastructure as a Product

Prioritize operational clarity, high-resolution observability, and zero-friction developer loops.