Arpit Bhayani: Asli Engineering

What is on the blog

As of September 2026 the blog lists 111 essays in five categories:

• Distributed systems (49 posts): replication, consensus, sharding, caching and the internals of widely used infrastructure
• Algorithms and data structures (24 posts): often the structures databases and systems rely on, explained through implementation
• Systems internals (21 posts): how storage engines, indexes and operating system features work
• AI systems (11 posts): LLM inference, embeddings and what AI-first engineering teams look like
• Career growth (6 posts)

Recent posts include pieces on engineering organisations in the AI era and the idea of "slop debt". The site also collects paper walkthroughs, video explanations, book recommendations and an RSS feed, and Arpit runs a newsletter and has built open source projects such as DiceDB. He publishes widely elsewhere too, with more than 210k YouTube subscribers according to his site.

Alongside the free writing he sells paid cohort courses: a System Design Masterclass, System Design for Beginners and an Applied AI Masterclass. Check his site for current dates and prices.

Why it is worth reading

Most system design writing stops at "shard the database" or "add a cache". These essays go a level down: how a specific system implements it, what breaks, and what the alternatives cost. That specificity is what turns memorised terms into understanding you can defend under questioning, whether in a design review or an interview.

How to use it

• Read by topic, not by date. Essays are independent. Pick the area you are weakest in and read three or four posts together.
• Implement something small. Many posts describe a mechanism you can build in an evening: a consistent hashing ring, a bloom filter, a simple LSM tree. Doing it once cements the idea.
• Follow the references. Posts often point at papers or source code; following one link per essay builds real depth.
• Use the videos when a concept is easier to see than read.

A post takes twenty to forty minutes to read properly.

The free archive or the paid masterclass?

The free archive is enough to learn a great deal of distributed systems and internals on your own. The paid courses add structure, live sessions and a cohort, which helps if you struggle to study consistently. Read a dozen free essays first; if the style suits you and you want guided practice, the paid courses are the natural next step.

Common questions

Is it beginner friendly? Some posts are, but most assume you know basic data structures and have built a backend. Beginners should start with a fundamentals course.

How is it different from ByteByteGo-style content? Fewer diagrams, more depth. It explains why and how, rather than giving a quick visual overview.

Is it only relevant to Indian or South Asian readers? No. The content is general engineering, though his audience is especially large in South Asia.

Best for: backend engineers who want to understand distributed systems and databases at the implementation level.