Our 2026 stack: the tools behind Cerebrix

People ask us what we use to build everything on this site, from the extensions to the docs. The honest answer is boring on purpose: a small, proven set of tools we know how to make sing.

Frontend

React and Next.js handle every web property we ship, with TypeScript everywhere and Tailwind for styling. The trade-offs are well documented and the ecosystem is huge — for a small collective, that velocity matters more than novelty.

Why TypeScript everywhere

Every line of product code we ship is typed. That sounds like a rule, but it's really a time-saver:

• Refactors that touch five files are still safe to do on a Friday
• Schemas and API responses share types, so drift is a compile error, not a bug report
• New members read the types to understand the system faster than any doc

TypeScript's typechecking cost used to be the argument against it. With the native compiler, that argument is gone — our CI typecheck step is now one of the fastest jobs in the pipeline.

Backend and data

• Node.js + Express or NestJS for APIs
• PostgreSQL as the default database, Redis for caching and queues
• Docker for local parity and predictable deploys

We pick PostgreSQL unless there's a concrete reason not to. It handles the 90% case — relational data, transactions, JSON when you need it — without forcing a second system into the picture.

Hosting and CI

Everything deploys through Vercel and GitHub Actions. The pipeline is deliberately boring: push to main, tests run, the preview builds, and a click ships it. For a small team, the least surprising deployment is the best one.

The rule that binds it

Every choice above passes one test: could a new member of the collective be productive with it by the end of their first week? If the answer isn't yes, we pick something else — no matter how interesting the alternative looks.

What we're watching

Three things get our attention when we evaluate new projects: edge runtimes for low-latency APIs, local-first sync for consumer tools, and whatever the TypeScript team ships next. We'll write about each as we form opinions — the blog is where those opinions land before they become stack decisions.