Notes on retrieval, embeddings, and the backend work behind them.
18 min readSep 16, 2026
A practical, technically grounded map of what separates calling an LLM API from real AI engineering: tokens, context windows, embeddings, RAG, agents, evaluation, security, and cost, explained for developers who already know how to build software.
6 min readSep 16, 2026
Array vs Set vs Map, Big O, binary search, and finding duplicates, explained with real production examples instead of textbook trivia.
13 min readSep 10, 2026
Most explanations of these concepts stop at the definition. Scope gets a rule about hoisting. Closures get a counter function that increments a number nobody…
12 min readSep 5, 2026
If you already know Express well, NestJS isn't a new backend paradigm, it's a way of organizing the things you already do. Here's a practical way to understand NestJS, mapped directly to Express, using Prisma.
7 min readSep 1, 2026
Every new project starts with the same authentication build: signup, verification, password reset, OAuth, protected routes. I got tired of building it from scratch, so I built it once, properly, and open sourced it.
10 min readAug 31, 2026
My in-memory cache worked perfectly locally and was silently empty on every request in production. Here's why Vercel's serverless model breaks caching, rate limiting, and background jobs, and what actually fixed it
8 min readAug 28, 2026
I added AI-powered lead classification to my form tool. Here's how I designed it to categorize, prioritize, and summarize incoming form submissions without putting AI on the critical path.
8 min readAug 25, 2026
Why your API might be making 101 database queries instead of 2. Learn what the N+1 query problem is and how to fix it with SQL joins and ORM eager loading.
9 min readAug 25, 2026
What actually separates senior engineers from beginners isn't the framework they know. Learn why understanding fundamentals, product thinking, and system design matters more than ever in the age of AI coding tools.
9 min readAug 21, 2026
What a retried request does to a money transfer, and the one column that makes retrying safe.
6 min readAug 20, 2026
A working model of vectors before the library magic, written while wiring the first retrieval path.
14 min readAug 20, 2026
Learn how to build an AI email auto-reply system using RAG, PostgreSQL, and pgvector, from embeddings to retrieval to LLM generated replies.
13 min readApr 22, 2026
Two sum, binary search, constraint validation, flattening nested data, and debouncing aren't just interview questions. Here's where each pattern actually shows up in real production TypeScript.
14 min readMay 30, 2024
React live coding interview questions
7 min readJun 19, 2019
A plain, honest explanation of what computer programming actually is, how programs get built, the tools engineers use, and what separates a good programmer from a mediocre one.