Hi, I'm Md Ashik
I write about backend engineering, system design, and the practical tradeoffs behind shipping software that holds up in production. Most of what I publish here started as a problem I had to solve at work.
Database round-trips were the bottleneck in our NestJS API. Here is how a disciplined cache-aside layer dropped P95 latency from 340ms to 28ms without making invalidation painful.
I have shipped REST, GraphQL, and now tRPC in production. For an app where one team owns both the frontend and the backend, tRPC is the most productive by a wide margin.
When I joined, the cluster was a single replica set handling 40K daily users. 18 months later we were at 200K with a 2x-improved P95. Here is what actually moved the needle.
Database round-trips were the bottleneck in our NestJS API. Here is how a disciplined cache-aside layer dropped P95 latency from 340ms to 28ms without making invalidation painful.
I have shipped REST, GraphQL, and now tRPC in production. For an app where one team owns both the frontend and the backend, tRPC is the most productive by a wide margin.
When I joined, the cluster was a single replica set handling 40K daily users. 18 months later we were at 200K with a 2x-improved P95. Here is what actually moved the needle.