How a radio station that runs itself works

NexRadio has no studio and no editorial team. What you hear is put together every hour by the NexRadio platform: the music selection, the presenter, the news, the weather, the station sound and this website. This is the chain, step by step.

The clock: a plan for every hour

Radio runs on an hourly clock. For every hour it is fixed where the news, the weather, the jingles and the talk go, and which kinds of music come in between: uptempo, calm, a classic, a request. NexRadio fills that clock automatically, hours ahead, following rules a programme director would set too: no artist twice within a few hours, no two ballads back to back, the right mood for the time of day.

Towards the end of the hour the planner counts backwards: whatever no longer fits is not scheduled, so the news starts on time without cutting a track short.

The music: a library that prepares itself

Every track that comes in is measured: tempo, key, energy, loudness, and the points where it may start and hand over to the next one. Those cue points decide how tight a transition sounds and where a presenter may talk over the intro without touching the vocals. Artwork, years and metadata are completed automatically.

The presenter: a voice that knows what is going on

NexRadio's presenter is an AI voice with a character of its own and its own show per daypart. Before each track the platform writes a short script based on what just played, what is coming, the time and the weather. That script is voiced, brought to broadcast loudness and placed on a music bed that keeps running until the next track starts.

The presenter only talks over the intro of a track, never over the vocals, and never over a news bulletin.

News and weather, fresh every hour

On the hour NexRadio brings a news bulletin followed by the weather for Ibiza. Both are written and voiced shortly before air, with their own newsreader and their own bed, so they sound like a real bulletin rather than a text being read out.

Requests with a shoutout

Listeners pick a track from the music library and add their name and town. The platform schedules the request in the current hour and the presenter announces it personally, usually within five minutes.

The station sound

All audio passes through one processing chain: transitions, jingles, loudness and the sound of the station are handled in one place, with the same processing big stations use. Jingles always play out; music is never turned down under a voice.

Stream, website and app

The broadcast leaves as 320 kbps MP3 and as HLS, 24 hours a day. This website, the listener app and the play archive are filled by the same platform: what you see here is what happens on air.

NL EN ES · See the playlist