harsh sharma

High-performance rate limiting service in Go using the token bucket algorithm.

• —Built a distributed rate limiter in Go using the token bucket algorithm, sustaining 10.8K+ req/sec with p99 latency under 36ms under concurrent load.
• —Implemented distributed rate limiting using Redis and Lua atomic scripts to ensure consistent token consumption and eliminate race conditions across application instances.
• —Designed a storage-agnostic architecture with Go interfaces, enabling zero-change backend swapping between in-memory and Redis stores; added IP-based limiting and rate-limit headers.

Lightweight full-text search engine in Go — indexes .txt files and serves queries over HTTP.

• —Inverted index built from scratch for fast and efficient text retrieval.
• —Lightweight HTTP server exposed for rapid search queries.

Interactive system design learning platform with visual explanations and hands-on challenges.

A minimal second-brain app to organize and save YouTube videos & Twitter links in one place.