Meet Maestrix, your marketing strategist

What are you marketing today?

Tell Maestrix about your product or your challenge. It plans the work and runs the skills, grounded in your brand.

Free, 10 credits, no card

Prefer to work in Claude, ChatGPT or Gemini? Browse free marketing skills

Used by fast-growing teams and prestigious universities

USC logoDTU logoBubble logoChainlink logoAurora logoCamunda logoUpsun logoPlanera logoGTM Nexus 360 logoKognitos logoSepteo logo

Strategy artifacts, not blog spam · personas, positioning, GTM plans, messaging frameworks

Built on leading Claude and Gemini models · connectors over MCP

One workspace, every brand and every task

The library, the tasks and the work in flight sit in the same place, so nothing you generate ends up outside the system.

Home

Good afternoon, Guillaume

Write a LinkedIn post about our new reporting feature…

Select a project
Value propositionsLinkedIn adsPage audit

Recent work

GTM plan · Acme
Persona review · Northwind
Ad angles · Contoso
Project overview

Acme

Brain 82% complete
Positioning
Buyer personas
Messaging
Value props

LinkedIn ads · Draft 2

Written from the Acme Brain: positioning, personas and proof applied automatically.

Saved to the library, ready to share or export

The real problem is context fragmentation

Strategy gets written in one tool, personas in another, ads in a third and the landing page somewhere else. Every hop loses the thread.

Most AI tools are document-centric. Prompt, generate, output disappears. The knowledge stays trapped in a chat nobody can reuse.

01

The output disappears with the chat

What you taught the model is gone when the tab closes. The next task, the next person and the next month all start from zero.

02

A strategy change never propagates

You sharpen the ICP in one document. Nothing generated last month knows, and nothing generated tomorrow inherits it either.

03

Every asset re-argues the strategy

Positioning, buyer language and proof get retyped into a prompt for the fifth time, slightly differently, and the work drifts.

The object library

Eleven object types per brand. Every task reads the subset it needs.

These are reusable objects, not a prompt and not a brand PDF. Edit a persona once and everything generated after it inherits the change.

Brand identity

Who they are and what they stand for

Product and insights

What it does, and what customers actually say

Ideal customer profile

Who they sell to, and who they don't

Buyer personas

The humans in the room

Problems and jobs to be done

The pain that makes them move

Value proposition ladder

Feature to capability to benefit to outcome

Positioning and competitors

The frame they can win in

Voice and tone

How everything should sound

Proof and evidence

What they are allowed to claim

Objections and risks

What stalls the deal

Channels and motion

Where the work gets published

Rulings sit alongside the pillars. Decisions you have already made, that later work is not allowed to contradict.

See all eleven pillars

Combinations, not one-off prompts

Sophisticated teams do not write every campaign from scratch. They keep a messaging architecture and generate assets from it. Maestrix stores that architecture, then combines it.

Persona × Use case × Value proposition

PersonaUse caseValue proposition
VP of SalesForecast accuracyReduce pipeline uncertainty
RevOps ManagerData consolidationOne source of truth
CRORevenue growthImprove win rates

Persona × Industry × Pain point

PersonaIndustry pressureAngle
Manufacturing CFOSupply chain volatilityCost control
Healthcare Ops DirectorStaffing shortagesEfficiency
Retail CMORising acquisition costMargin protection

Change the ICP once and every asset generated after it inherits the update. That is the part a prompt library cannot do.

Marketing AI needed an operating system, not another content generator

Personas, value propositions, use cases, objections, competitors and proof stored as reusable objects instead of prompts. Every campaign, ad, email and page is generated from them.

11

reusable object types per brand

6

guardrails on every output

Unlimited

brands, kept separate

Structured, so it can be queried

Ask your marketing, not your chat window

Personas, pains, use cases, features, value propositions, proof and assets are linked to each other. That turns a content generator into an operating system for strategy.

Show every campaign targeting CFOs worried about compliance.

Generate LinkedIn ads for every manufacturing persona using use case three.

Which value propositions are underrepresented in our content library?

Build an ABM campaign for the personas matching our highest-converting segment.

Three ways to seed a brain

Pick one, mix all three. Nothing is locked after setup.

from a URL

Generate it

Give a client name and a website. Maestrix drafts all eleven pillars, you fix what is wrong, you save.

Six guardrails on every output

Applied silently to every generation, so your time goes on judgement calls instead of deleting "seamless" and "game-changing".

checking output · 0/6
✓

Truth floor

No invented facts, metrics, logos or quotes

✓

Specificity

Concrete nouns beat adjectives

✓

Buyer-first framing

Their problem before your product

✓

Voice fidelity

Sounds like the brand, not like an LLM

✓

One decision

Every asset asks for exactly one next step

✓

Ship-ready

Paste it and go

Every output inherits, and the library gets better

Five steps. Each one leaves the brain better than it found it.

01

Seed the brain

Generate from a URL, distil from documents, or paste raw text. Everything stays editable.

02

Set the rulings

Write down the decisions already made. Later work respects them without being reminded.

03

Run a task

Each task carries a role, an objective, an output structure and a quality bar, and pulls the pillars it needs.

04

Sharpen and chain

Every claim in the result is graded against your sources — Documented, Provisional or Inferred — then the output and its context hand off to the next task.

05

Save it back

Approved work writes into the brain. The next asset starts from a better brief than the last one did.

Pro only

One brand or fifty, without the libraries bleeding

Lite runs one brand well. Pro is for product lines, portfolios and agencies running several at once.

  • Real separation between brands

    Each brand gets its own objects, its own rulings and its own voice. Nothing crosses over, including in chat.

  • Agents that chain tasks into workflows

    Approve step by step, or run an approved sequence and collect the output. Batch a task across several brands at once.

  • Scheduled work and stakeholder reporting

    Daily, weekly or monthly runs with digest notifications, and reporting a stakeholder can read without translation.

  • Two competitive workflows, not one

    Competitive Analysis finds who shapes a market and how to compete in it, no URLs needed. Competitor Scan tears down one named company.

Your brands · pick one

Switching brands switches every pillar, ruling and writing style behind the work.

Strategy artifacts, not blog posts

Over 100 tasks that return the things marketing actually runs on. Every one states what you get, roughly how long it takes and what it costs, before you spend a credit.

Browse the full catalog by discipline
~3 min · 1 credit

ICP definition

Sharpen who to chase, and write down who to ignore

You get: Scored ICP with anti-ICP

~3 min · 1 credit

Buyer persona profile

A persona with a job, a fear and a calendar, not a stock photo

You get: Persona dossier

~3 min · 1 credit

Positioning canvas

Pick the frame you can win in, and the ground you concede

You get: Canvas and statement

~3 min · 1 credit

Competitor battlecard

How you win, lose and reframe against a named alternative

You get: Sales-ready battlecard

~4 min · 1 credit

GTM strategy brief

A defensible plan built from the brain, not from a template

You get: 6-section brief and 90-day plan

Claude · maestrix MCP

you › Draft a LinkedIn post for Northwind in our voice.

⚙ maestrix.get_brain · positioning, voice, competitors

⚙ maestrix.run_task · linkedin-post

⚙ maestrix.save_task_result · Northwind Robotics

✓ Saved to the right project.

Connectors ready out of the box. Any MCP server with a URL can be added.

SemrushAhrefsSearch ConsoleHubSpotSalesforceApolloG2IntercomGoogle AdsLinkedIn AdsMeta AdsWebflowWordPressNotionGoogle DriveSlackBufferFigmaCanvaLinearAsanaAirtablePostHogGA4Zapier
Claude

Your library travels to the tools you already work in

Maestrix connects over MCP, so Claude can read a brand's objects, run your tasks and save approved work back into the right project.

Positioning, voice and competitors loaded before a single word gets written.

Set up the MCP server

Done with you

Some teams want the brain built for them

Our strategists build it, run the first campaigns with your team in the room, and hand over a workspace you keep running yourself.

1

Brain build

Two weeks. We seed every pillar from your material, pressure-test the positioning and leave you a working brain.

2

Pod sessions

Your team and ours in the same room, shipping real campaigns while learning the workflow on live accounts.

3

Custom agents

Workflows wired to your stack, built once and left behind as reusable agents your team owns.

Engagements from $3,000 a month. Same product, same brain format, yours to keep.

Who it is for

Same engine underneath. Pick by how many brands you market and who you report to.

Solo operator

Founder, freelancer, fractional marketer

Start free

One brand, one owner, no seats to configure. Get a real brand brain and start shipping the same afternoon.

  • One brand brain, 11 pillars
  • Full task catalog
  • MCP connectors
  • 10 free credits, no card

In-house team

Marketing team at a multi-product company

Maestrix Pro

Several brands and several logins. Roles, scheduling and a shared brain so nobody re-explains the company to the model.

  • Unlimited brands
  • Team seats and roles
  • Scheduling and recurring tasks
  • $200 / month

Agency

Client work, reporting, delivery at volume

Maestrix Pro or Sprints

Client separation and reporting are the whole job. Run it yourself in Pro, or have our strategists deliver a fixed-scope sprint.

  • Client workspaces
  • Shareable client reporting
  • Agency workflows and pods
  • Sprints quoted per engagement

Not sure which? See pricing

Priced by output, not by seat

One credit runs one task. Exports, brain edits and presets are free.

Start free0credits · no card

The full product. Run your first tasks on your own brand.

Top-up0credits · $55

A full positioning and persona foundation, with room to iterate.

Top-up0credits · $99

A monthly content and campaign cycle for one brand.

Top-up0credits · $185

A month of campaigns across several brands, with scheduled work.

See plans and what a credit buys

Questions, answered

No, and the difference is the brain rather than the model. A general chat tool starts every request with no idea who your client is, so you supply the context by hand and it is gone when the tab closes. Maestrix stores that context as eleven structured pillars per brand, feeds each task the subset it needs, and applies six guardrails to the result. The model does the writing. The brain decides what is true.

Your first marketing brain takes about ten minutes

Add your website, review the eleven object types, fix what is wrong, then run a task.

10 credits free. No card. Paid credits accumulate and do not expire.