All guides, page 10 of 21

Newest first. Search or pick a topic.

482 guides, newest first

• CAP Theorem vs PACELC: The Latency Half CAP Leaves OutSystem design · A plain-English guide to the CAP theorem and PACELC, showing what CAP says about network splits and the everyday latency trade-off it leaves out, built on one running homestay booking example.25 min read · Updated 22 Oct 2026
• How a CDN Works: Push vs Pull, Cache Keys and PurgesSystem design · A plain-English guide to how a CDN works: edge servers, push and pull, cache keys, TTLs, purges and a test edge you can run, shown on one running example of a streaming app's premiere night.28 min read · Updated 22 Oct 2026
• Circuit Breaker Pattern: Fail Fast When a Dependency Is SickSystem design · A plain-English guide to the circuit breaker pattern: its three states, trip rules, fallbacks and retries, shown on one running example of an online shop's payment gateway.20 min read · Updated 22 Oct 2026
• Coding Agent Architecture: Harness, Tools and SandboxesAI and GenAI · A plain-English guide to coding agent architecture: the harness loop, tools, sandboxes, context on big repositories and test checks, shown on one running example of a billing bug fix.24 min read · Updated 22 Oct 2026
• Consistency Models: From Linearizable to EventualSystem design · A plain-English guide to consistency models: linearizable to eventual, each step shown with a real system and one feature of a live sports app.26 min read · Updated 22 Oct 2026
• Consistent Hashing Explained: Add Nodes Without a ReshuffleSystem design · A plain-English guide to consistent hashing, from the hash mod N trap to rings, virtual nodes and bounded loads, built on one running image cache example.24 min read · Updated 22 Oct 2026
• Debugging AI Agents: Reading Transcripts and Stop ReasonsAI and GenAI · A plain-English guide to debugging an AI agent that loops or stops early: how to read a transcript round by round, what stop reasons mean, the common loop bugs and what to log, shown on one running IT helpdesk example.22 min read · Updated 22 Oct 2026
• Leader Election, Leases and Distributed LocksSystem design · A plain-English guide to distributed locks, leases, fencing tokens and leader election with Redis, ZooKeeper and etcd, built on one running warehouse software example.26 min read · Updated 22 Oct 2026
• Design a Distributed Task Scheduler: Cron at ScaleSystem design · A plain-English guide to designing a distributed task scheduler: due-job polling, row locks, leases, duplicate checks, retries and priorities, shown on one running example of a bill and EMI reminder service.26 min read · Updated 22 Oct 2026
• Delivery Guarantees: At-Most, At-Least and Exactly-OnceSystem design · A plain-English guide to at-most-once, at-least-once and exactly-once delivery, and why exactly-once is really effectively-once, built on one running loyalty points example.22 min read · Updated 22 Oct 2026
• Quadtrees and Geohashes: How Apps Find Nearby Riders FastSystem design · A plain-English guide to spatial indexing with geohashes and quadtrees, built on one running example: a Pune bike-taxi app finding the nearest free rider.24 min read · Updated 22 Oct 2026
• How LLMs Are Trained: Pretraining, Fine-Tuning and RLHFAI and GenAI · A plain-English guide to every stage of LLM training, from cleaning data and pretraining to instruction tuning, RLHF, DPO, safety tuning and testing, shown on one running example of a farmers' helpline model.25 min read · Updated 22 Oct 2026
• How LLMs Work: The Life of a Prompt, Step by StepAI and GenAI · A plain-English walk through what happens inside a large language model, from tokens and embeddings to attention, next-token probabilities and the context window, built on one running recipe-app example.26 min read · Updated 22 Oct 2026
• How to Evaluate AI Agents: Trajectories, Not Just AnswersAI and GenAI · A plain-English guide to testing AI agents: check the real end state, grade every step, pick graders and run each task many times, shown on one running example of a shop's ordering agent.24 min read · Updated 22 Oct 2026
• Hybrid Search: How BM25 and Vector Search Work TogetherAI and GenAI · A plain-English guide to hybrid search: how BM25 keyword search and vector search each fail, how to merge them with RRF or weighted scores, and how to measure the result, shown on one running auto parts store example.25 min read · Updated 22 Oct 2026
• Idempotency Keys Explained: Safe Retries Without Double ChargesBackend · A plain-English guide to idempotency keys: how they make API retries safe and stop double charges, built on one running UPI bill-payments example.22 min read · Updated 22 Oct 2026
• Design a Key-Value Store: A Dynamo-Style GuideSystem design · A plain-English guide to key-value store design in the Dynamo style: partitioning, copies, quorums, conflicts and repair, shown on one running example of an online marketplace's shopping carts.26 min read · Updated 22 Oct 2026
• Lamport Clocks and Vector Clocks: Ordering Without a ClockSystem design · A plain-English guide to Lamport clocks and vector clocks: how devices order events without a shared clock, spot concurrent edits and merge them. It uses one running example of an offline notes app.25 min read · Updated 22 Oct 2026
• LLM Structured Output: JSON Schemas That Don't BreakAI and GenAI · A plain-English guide to getting reliable JSON from a model, from JSON Schema and constrained decoding to validation, retries, missing fields, versioning and tests, shown on one running example of a GST invoice reader.25 min read · Updated 22 Oct 2026
• Temperature, Top-p and Top-k: LLM Sampling ExplainedAI and GenAI · A plain-English guide to LLM sampling: how temperature, top-k, top-p and seeds change each next-token pick, and why vLLM and llama.cpp apply them in a different order, shown on one running example of a support bot and a tagline writer sharing one model.28 min read · Updated 22 Oct 2026
• LLM Tokenization Explained: Why Hindi Costs More TokensAI and GenAI · A plain-English guide to LLM tokenization: what tokens are, how byte-pair encoding works, why Hindi needs more tokens and how to count them before you send, shown on one running Hindi news app example.25 min read · Updated 22 Oct 2026
• Load Balancers Explained: L4 vs L7, Global vs LocalSystem design · A plain-English guide to load balancers: L4 vs L7, algorithms, health checks, sticky sessions and global routing, with nginx and HAProxy behaviour we tested, shown on one running ticketing site example.30 min read · Updated 22 Oct 2026
• ML System Design Interview Framework: Six StepsAI and GenAI · A plain-English guide to the ML system design interview: six steps, a 45-minute plan, a position-bias check and a practice rubric, shown on one running example of e-commerce search ranking.28 min read · Updated 22 Oct 2026
• Monolith vs Microservices: When to Split, and When Not ToBackend · A plain-English guide to monolith vs microservices: what each really costs, when a split pays and how to split safely, shown on one running example of a growing matrimony site.23 min read · Updated 22 Oct 2026