study.jaspreet.info

Notes worth
keeping.

Long-form notes on the things a backend engineer is asked to know cold — written to be read once in order and then used as a reference. Mechanisms drawn first, then the code that shows them.

Ten pages

The Backend Ladder

Systems thinking that survives contact with production, the languages you will be asked to read, and the cloud those systems actually run on. Every page prints properly — Ctrl-P gives a paginated document with the code wrapped rather than clipped.

The Backend Ladder Start here · overview of all three pillars The map: what each mechanism is, sorted by depth, with a route through the other eight pages. The Machine Room Deep dive · foundations Latency, Little’s Law, the utilisation knee, delivery semantics, cache arithmetic, backpressure and metastable failure. Reading Go Deep dive · Go What a slice really is, where an allocation comes from, which timeout the server is missing, and why that goroutine is still running. From Account to Pod Deep dive · cloud Every hop between a user typing a hostname and your handler running — who authorised it, what it costs, what breaks it. Python, End to End Foundations · Python Names that are not variables, defaults evaluated once, a lock that serialises your threads, and an object model that costs bytes. Java, Then Spring Tutorial · Java + Spring Boot A compiler that actually says no, a runtime tuned for twenty-five years, and the framework that assembles the rest. AWS, Service by Service Deep dive · AWS + SDKs What each service does when your code calls it — concepts, defaults, failure modes and working code, in Go and Python. Data Structures, In Motion Deep dive · algorithms, interactive Every structure and algorithm worth knowing, with figures that run — step the partition, the frontier, the table — and code in Go, Python and Java. Prompt to Production Deep dive · AI engineering, interactive What an AI engineer is actually hired to know — tokens, retrieval, agents, evaluation, serving and cost — with fifteen figures you can drag, in Python, TypeScript and Go. Everything They Ask Map · the whole technical interview Thirteen areas and forty-six topics, from the shape of the loop to the last five minutes — with the sub-topics under each spelled out rather than implied.

Also in the repository

Eight worked system-design problems — news feed, URL shortener, WhatsApp, typeahead, Dropbox, Uber, rate limiter, YouTube — live in system-design/ as Markdown, and are read on GitHub rather than served here.