Category
10 articles on system design
A system design interview walkthrough for a production RAG and AI agent orchestration platform — the retrieval-augment-generate loop, tool-call failure containment, hallucination guardrails, and why prompt injection is a first-class threat model here.
A system design interview walkthrough for a multi-channel notification system handling billions of push, email, and SMS sends a day — async fanout, per-channel queues, dedup, and transactional-vs-bulk prioritization.
A system design interview walkthrough for an Uber-style dispatch system matching riders to nearby drivers at scale — geospatial indexing, batched matching, real-time location ingestion, and the surge-pricing feedback loop.
A system design interview walkthrough for a real-time messaging system handling 2 billion messages/day — persistent connections, offline delivery, message ordering, and multi-device sync, WhatsApp-style.
A system design interview walkthrough for a distributed rate limiter enforcing per-user and per-endpoint limits across 1M requests/second — token bucket vs. sliding window, sharded Redis counters, and the thundering-herd problem at window boundaries.
A system design interview walkthrough for a multi-tenant LLM inference platform serving 50,000 requests/second — continuous batching, KV-cache economics, and prefill/decode disaggregation.
How to build a system handling 10,000 ad clicks per second — Kafka for event queuing, Flink for stream processing, deduplication, and ClickHouse for real-time analytics.
High-level design of a live streaming system at FIFA scale — CDN architecture, adaptive bitrate, low-latency delivery, and where things actually break at 50M viewers.
A system design interview walkthrough for a ranked social feed at 500M daily active users — fanout-on-write vs. fanout-on-read, the celebrity-account problem, and how ranking actually happens at serve time.
