Node.js is one of the environments I utilize most frequently in applications when an endpoint or API is needed (unless I opt for python, which typically would only be chosen if the task calls for certain requirements like machine learning). I have deployed many express applications with REST, GraphQL, and work-queue driven or imperative APIs exposed to other services. Typically, statelessness is a driving goal wherever possible, and I prefer JWTs alongside clean, versioned routing conventions.

Today I typically begin any new node project with typescript, as I do on the front-end, and am adept at writing intelligent unit tests.

Other uses of Node.js in my prior projects include isomorphic apps with server-side rendering in React, building ORM-like tools to move between massive university relational databases and clean objects, and proxying or wrapping other services like elasticsearch for exposed APIs. I have also developed web assembly code, typically in rust, to run equivalently and efficiently as part of an isomorphic pipeline for the client and Node.js on the server.

Working at Slite for 4 years, I maintained a complex Node.js stack for distributed, live collaborative editing backed by Operational Transform and ShareDB, as well as working with a large Node.js-based backend for an application with high volume and feature coverage.