Techno Music World is a live content platform for the global techno scene, organised into six
sections — events, artists, music, news, articles and a glossary — each with its own route tree
and all of them published from Sanity.I built this as a contract developer for Cyber Koala, working across the stack — the Next.js
frontend for all six sections, the API surface, and the Sanity content architecture behind them.
The codebase was shared rather than mine alone, and this page is scoped to my part of it.TODO(razvan): give the date window. "Built between X and Y" is a fact you have and I do not — it
is the last thing missing from this opening.
What was verified, and what was not
The technical description on this page was read out of the served HTML of technomusicworld.com on
2026-09-16 — Next.js build output, 394 references to the Sanity CDN, Tailwind utility classes with
custom theme tokens, self-hosted Atiga display faces, and a JSON-LD layer emitting MusicGroup,
MusicRecording, ItemList, SearchAction, Organization and WebSite.That evidence establishes what the platform is. It establishes nothing about who wrote which
part of it, and there is no repository reachable from this machine to measure. The entry therefore
carries provenance: "deployed", whose own type documentation warns that authorship under it is
asserted rather than measured.TODO(razvan): if you can get the repo onto this machine, the ownership can be measured the way
OWNERSHIP.md measures everything else, and this section gets deleted in favour of a real commit
count. That is the version worth having.
Content architecture
TODO(razvan): six sections sharing one Sanity dataset is the interesting structural problem here —
how the schemas relate, what an artist document has to carry so that it can be referenced from an
event and a release, and how the routes are generated. Write the parts you worked on.
Structured data
The JSON-LD is the most distinctive thing visible from outside: the pages emit music-domain
schema.org types rather than generic Article markup, which is what makes artists and releases
legible to search engines as music entities.This was part of my scope. TODO(razvan): say which types you modelled and what each one solved —
why an artist is a MusicGroup rather than a Person, what ItemList buys on a listing page,
and what SearchAction exposes. This is the most distinctive engineering on the site.
Challenges and Learnings
TODO(razvan): two or three, one H3 each, scoped to your own work. The same framing note from the
newsletter case study applies — "a specific problem I owned inside a larger system" is a stronger
and more honest story than an implied claim on the whole platform.
Outcome
TODO(razvan): no outcome numbers are established for this project. Do not invent traffic, ranking
or engagement figures. If you have real ones and they are publishable, put them here; if not,
describe the outcome in shape rather than magnitude.
Related projects
Multi-Brand Content & AI Platform
A multi-domain publishing platform combining server-rendered content, headless CMS workflows, technical SEO and an authenticated streaming AI assistant.
A React Native bedtime-stories product on iOS and Android: parents generate a personalised, illustrated story for their child and have it narrated in one of four voices.
