A small set of projects that demonstrate how I think about performance, resilience, and operating systems in production.
Handles tens of thousands of requests per second with consistent throttling guarantees.
Distributed sliding-window rate limiter designed for high-throughput APIs. Focuses on predictable behavior under burst traffic and failure scenarios.
Database time machine — captures every change from Postgres and lets you replay history safely.
Change Data Capture system that streams PostgreSQL WAL changes into Redis Streams for low-latency consumers and Kafka for durable archival. Supports controlled replay of historical events via gRPC and a Django admin dashboard, so you can debug, audit, or rebuild downstream state without touching production traffic.
Single pane of glass for latency, error rates, and saturation across services.
Lightweight monitoring dashboard that pulls metrics from Prometheus and traces from OpenTelemetry to help reason about incidents.