Skip to content
My Primers

System Design Primer

Deep technical primers for distributed systems, LLM infrastructure, and engineering interviews.
Start with Primers LLM System Design
LLM Primers Long-context LLMs, quantization, speculative decoding, attention variants, and context parallelism.
LLM System Design Context management, KV caching, inference optimization, agentic workflows, distributed training, and vector DBs.
Diffusion Diffusion and world-model interview map: DDPM, solvers, tokenizers, DiT, flow matching, and AV architectures.
Distributed Systems Primers Partitioning, replication, consensus, storage engines, caching, stream processing, analytical systems, and RPC.
System Design Rate limiters, distributed counters, leader election, object stores, timeseries DBs, feature stores, and ML infra.
Misc Performance debugging and other miscellaneous topics.
Linear Algebra & Backprop Two-layer classifiers, MLPs, attention, FFN, KL divergence, and matrix calculus — from scratch.
Modern Coding KV stores, in-memory spreadsheets, credit ledgers, web crawlers, serialization, and more.
40-min Interviews Rate limiters, distributed counters, caching, job schedulers, pub/sub, and notification systems.
PracHub Designs Deduplication, hosted notebooks, frequency capping, priority schedulers, and S3-style storage.