Projects

A selection of backend and distributed systems projects I’ve built to explore reliability, performance, and real-world constraints.

Distributed background job queue CLI.

• Worker-pool based execution with bounded queues and back-pressure.
• At-least-once delivery with retries, exponential backoff, and idempotency guarantees.
• Graceful shutdown: drains in-flight jobs and exposes health endpoints for orchestration.

Go · CLI · Concurrency · Context cancellation · Docker

Log-structured distributed key-value store.

• Append-only storage engine with segment compaction and tombstones.
• Read path optimized with in-memory indexes and on-disk SSTable-style files.
• Explored locking strategies and concurrency control for multi-goroutine access.

Go · Storage engines · Concurrency · File I/O

Scalable request routing & rate-limiting layer.

• Fan-out/fan-in pattern to multiple model backends with per-tenant limits.
• Token-bucket based rate limiting and circuit breaking for unhealthy upstreams.
• Caching hot responses and aggregating metrics for SLO monitoring.

Go/C++ · Reverse proxy · Rate limiting · Load balancing

Efficient intrusion detection for CAN bus traffic.

Python · ML · Networking · Linux