Content Engine

Content that learns.

You bring the expertise. We augment it. Personalised content at the scale your segments demand, strapped to your data, not pretending to be magic.

  • Tone-of-voice extracted from your existing site
  • Drafts, images and auto-publishing
  • Multi-LLM grading against your playbook
Most AI content tools flood you with content you have to fix.

Generic input becomes generic output becomes generic results. The marketer's brand voice, judgement and strategy are what make content work. They shouldn't be replaced. We do the writing. You bring the strategy.

Why Content Engine

Collaborator, not replacement.

AI tools that pretend to be magic age badly. Tools that respect human expertise don’t.

Content Engine reads your existing site, learns your tone, and writes against the playbook you set. The marketer brings strategy and judgement; we bring the writing, and we score every draft against your tone before you ever read it.

No fix-it-yourself slop. No content you didn’t ask for. The system gets better the more you tell it where it was wrong.

Where Content Engine plugs in:

Elevate: defines the segments the content is written for.

Close the Gap: measures the response so the engine learns what worked.

The Constant Media: for bespoke build of the publishing layer when self-serve isn’t enough.

Know your audience. Write for them. Measure the response.

What you get

Personalised content. At the scale your segments demand.

Drafts, images, publishing, and a learning loop that sharpens over time.

Research-backed drafts

Each piece written against a brief: topic, intent, audience, source links.

Auto-generated images

Visuals to match each draft, generated to your brand palette.

Auto-publish to your CMS

WordPress and Strapi today, with more on the roadmap. Published on a schedule or saved as drafts.

Topic suggestions from data

Engagement data (PostHog, GA4) tells the engine what is working. It writes more of that.

1–5 star tone review

Score each draft on tone. The system learns what landed and what did not.

Tone-of-voice protection

Multiple LLMs grade each draft against your playbook and customer bible before it reaches you.

How it works

Six steps. The expertise stays yours.

  1. Stage 1

    Learning

    We learn your tone from your existing site.

  2. Stage 2

    Write the rules

    You shape the playbook: topics, formats, frequency

  3. Stage 3

    Drafting

    We draft against the playbook with research, images and scheduling.

  4. Stage 4

    Grading

    Multiple LLMs grade each draft against your tone, playbook and customer bible

  5. Stage 5

    Review

    Score the draft 1–5 on tone, and the system learns

  6. Stage 6

    Published

    Auto-publish to WordPress or Strapi (with more on roadmap), scheduled or as drafts

Under the hood

AI tools that pretend to be magic age badly. Tools that respect human expertise don’t.

Engagement-data integration with PostHog and GA4 feeds the topic engine. Multi-LLM grading against your playbook keeps the brand voice intact. Tone extraction from your existing content means the system starts knowing what you sound like.

The marketer’s expertise is the input. The engine is the multiplier. The grade is the guardrail.

What you get
  • PostHog / GA4 engagement integration.
  • Multi-LLM grading against playbooks.
  • Tone extraction from existing content.

Your voice. Scaled.

A pilot starts with us reading your existing content, building the tone playbook, and showing you a sample piece in your voice. Twenty minutes of your time. Useful even if you don’t buy.

Pricing

Plans for every content team

Start small, scale as you publish. All plans pull live from Stripe.

Enterprise
Bespoke plan for in-house brand teams.
Call us
Talk to sales
  • Custom credit allotment
  • Dedicated success manager
  • SSO + audit logs
The Signal

A fortnightly note for marketers who would rather know than guess.

No spam. Unsubscribe in one click.

Content Engine — AI content that learns your voice | EMI