A planner agent owns the entire production process. It decides which specialists run, in what order, and what actually gets shipped, and every one of those decisions is visible and logged.
The critic agent reviews everything against your actual brand voice, tone, and the brief itself. Any disagreement gets surfaced, every rejection is logged, and a human always keeps the final veto.
Every specialist only touches the real world through a specific tool, whether that is generating something, rendering it, scheduling it, or publishing it. Each tool has clear limits, and every call it makes is traceable.
Performance data feeds back into the brief itself, not into the model. That means the system learns what is actually working at a level a human can read, understand, and override.
We sit with your actual marketing team and follow both the assets that ship and the ones that quietly never do. Most of how MERIDIAN ends up shaped comes directly out of this.
Six specialists with named roles, covering ingest, brief, write, render, critique, and publish, plus a planner that owns the whole production graph. All of it written down before any of it gets built.
A cloud native backend, video generation pipelines, real Meta API integrations, and attribution feedback built in. Observability from day one, not bolted on later.
A marketing operator owns the loop. They can read every brief, override any output, retighten the critic, and roll back any asset.
No. MERIDIAN replaces the production work, the actual mechanical labor of turning a brief into a set of assets shaped for each channel. The interesting work, the strategy, the brief writing, the judgment on what is even worth saying, stays completely human.
Teams we have built this for end up spending more time on strategy after it launches, not less. The system just gives them their attention back.
The critic agent enforces it. We start by actually writing your brand voice rules down in a form the system can check against, and the critic rejects anything that drifts from it. Drift is the most common way this kind of system fails, and the critic exists specifically to catch it.
If your brand voice genuinely cannot be written down in any consistent way, that is a different problem, and we will tell you honestly if that is what we are seeing.
Performance data gets summarized regularly and proposed to the human operator as changes to the brief, never as some silent retraining happening behind the scenes. The operator approves or edits the proposal, the brief gets updated, and the pipeline starts producing against the new version. The model never quietly changes its mind about your brand on its own.
The Meta family of apps, X, LinkedIn, and a generic webhook for whatever else your stack uses, plus email and CRM integration through API. Adding a new channel is mostly a matter of building the specialist for it, the planner picks it up automatically once that exists.
It varies a lot by engagement, and honestly we do not optimize for a number here. Quality through the critic is the actual goal, throughput is just whatever that quality allows.
Tell us what you are dealing with. We read through your existing workflow, get back to you within 48 hours with an honest architecture sketch, and go from there together.
Tell us what you are dealing with. We respond within 48 hours with an honest read and an architecture sketch.