Daily Neon — A Marketing OS That Runs on Agents

Ours, and the proof. Daily Neon is a product we own. It's also where our own agents do real work: every week, scheduled agents draft posts, briefs, blogs, and comics into it and wait for one-tap approval. The studio runs its own marketing on it.

What it does

A small business uploads the content it likes into a swipe file, sets up a brand voice and profile, and Daily Neon turns that into a steady stream of marketing content — social posts, blog drafts, weekly industry briefs, even a Friday comic — generated fresh and queued for approval. Think "the thing actually does the work, you just say yes."

Nothing auto-publishes. Everything lands as a pending draft the owner approves with one tap. That single design decision — generate freely, gate on human approval — is what makes an unattended content engine safe to point at a real brand.

How it's built

Rails 8 on the backend, with a deliberately wide agent and tool surface: drafts, a hopper, brand-voice modeling, scheduling, blog and comic generation, swipe-file analysis, and metrics — all exposed as tools an agent can call. A web app and a mobile app sit on top for review and approval.

The part that's hard to fake is the scheduling layer. Daily Neon isn't a button you press to generate one post. It's a set of recurring jobs: fill next week's calendar, draft follow-ups to whatever performed, write the weekly brief, refill the hopper. Those jobs run whether or not anyone is watching, and report back when they need a human.

Why it's the keystone

Most agencies can show you AI features. Very few can show you their own marketing being run by their own agents against their own platform, on a schedule, in production. Daily Neon is that. When we say a system can run unattended and only surface to a human when it matters, this is the thing we're pointing at.

See the lab writeups for how the agent fleet behind it is wired.

Want a content engine that runs itself? Start here.