How it works

12 AI agents. 5 human approval gates. Nothing ships without a person signing off.

Twelve specialized agents run the research, writing, technical checks, and publishing. But at five fixed points in that pipeline, a person has to review and approve the work before it moves forward — nothing reaches your site, or your inbox, without a human sign-off.

// 12 agents · 5 human gates · nothing ships unreviewed

The full pipeline

Agent by agent, gate by gate

This is the same sequence for every client. Strategy clients see it stop after Gate 2; Full Execution clients see it run all the way through to a published page.

Agent 01

Intelligence Agent

Pulls together your site content, competitor citations, and the prompt landscape relevant to your category.

Agent 02

Visibility Agent

Runs your tracked prompts against ChatGPT, Perplexity, Gemini, and Google AI Overviews, and logs exactly where you're cited, missed, or mentioned without credit.

Agent 03

Strategy Agent

Turns the visibility findings into a prioritized 30/60/90-day roadmap.

✓ Gate 1 — Roadmap approval

A strategist signs off before you see it

A human strategist reviews the roadmap for accuracy and priority before it ever reaches you.

Agent 04

Research Agent

Digs into the specific topics and queries the roadmap flagged as highest priority.

Agent 05

Content Brief Agent

Turns that research into a structured brief — entities to cover, direct-answer requirements, internal links, and target queries.

✓ Gate 2 — Content brief approval

Checked against the roadmap before anyone writes a word

A strategist checks the brief against the roadmap before it's handed to your team, or to our Writing Agent.

Agent 06

Writing Agent

Drafts the page from the approved brief. Full Execution clients only, from here on.

Agent 07

Brand Language & Compliance Agent

Scores the draft on voice, terminology, legal and claims language, and formatting against your brand rules.

✓ Gate 3 — Brand voice approval

Reviewed before a single line of code is written

A strategist reviews the draft and its compliance score before development starts.

Agent 08

SEO Technical Agent

Checks meta tags, heading structure, internal linking, structured data, AI-crawler accessibility, and passage-level citability before the page gets built.

Agent 09

Development Agent

Builds the approved page into your CMS or codebase on a staging preview.

Agent 10

QA Agent

Tests the staged page — links, schema, mobile rendering, load speed — against a fixed checklist.

✓ Gate 4 — QA approval

A person reviews the preview, not just the report

Someone on our team reviews the staging preview after the QA Agent's checks pass — not just the automated report.

✓ Gate 5 — Pre-publish approval

You give the final sign-off

You, or your named approver, sign off on the staging preview before anything is allowed to go live.

Agent 11

Publishing Agent

Pushes the approved page live and confirms it resolves correctly in production.

Agent 12

Experiment Agent

Tracks the published page's citation performance over time and feeds the results back to the Intelligence Agent for the next cycle.

Why this matters

Agents for consistency and speed. Humans for judgment and accountability.

The agents run the same checklist every time, at a pace no team of strategists could match on their own. The gates exist because a checklist can't take responsibility for a brand claim, a legal risk, or a judgment call about what's actually worth publishing — a person can, and does, at five separate points before anything reaches your site.

Most "AI visibility" vendors sit at one of two extremes: fully manual, which is slow and doesn't scale past a handful of pages a month, or fully automated, which ships whatever the model produced with no review. We built this to be neither — agents for the parts that benefit from speed and consistency, people for the parts that require judgment.

Bottom line: nothing an agent produces reaches your site, or your inbox, without a person having looked at it first.
Go deeper

Related reading

Program

Strategy

The audit, roadmap, and content briefs — stops after Gate 2.

See the program →
Program

Full Execution

We write, build, and publish — runs the whole pipeline through Gate 5.

See the program →
Proof

Methodology

Which engines we track, what counts as a citation, and how often we check.

Read the methodology →
Get started

See which tier fits your team

Strategy stops after Gate 2. Full Execution runs the whole pipeline, gate by gate, all the way to a published page.

Book a call