The autonomous operator
for media buying.

Kerdixo is the AI-native operator for affiliate media buying. It owns the decisions the ad platforms cannot make, and backs each one with a receipt.

At a glance
2026
Founded
United States
Headquarters
Private beta
Current stage
Meta · TikTok
Platforms at launch
Receipts
Per material decision

The ad platforms got smarter inside the campaign. Nothing got smarter around it.

Meta and TikTok optimize delivery inside a live campaign. They cannot rank offers, move budget across a portfolio, supply compliant creative, or call a lifecycle decision.

Kerdixo is the operator built for that surface. It compounds every measured outcome into a data advantage that cannot be bought, only accumulated.

A new category, stated plainly: the autonomous operating layer for media buying. Built in public.

The principles we build on.

  1. 01

    An operator, not a tool

    An operator that ranks offers, launches campaigns, refreshes creative, and shifts budget inside guardrails.

    The operator queue →
  2. 02

    Proof in public

    The metrics that earn trust go public. Customer data and the operating recipe stay private.

    The thesis →
  3. 03

    Guardrails before autonomy

    Action stays inside approved boundaries. The safety layer does not negotiate.

    Guardrails explained →
  4. 04

    Compounding, not noise

    Every outcome sharpens the next decision. The data history cannot be bought.

    The compounding loop →

Agent-first, and built to compound.

Kerdixo starts with the operator as the unit of work. The operator is both the product and the operating model, the structure that lets the company compound, not just grow.

Agent-first

The operator is the company, not a feature.

Most products bolt an assistant onto a human workflow. Kerdixo inverts it: the operator is the unit of work, and people set direction and guardrails. Output scales with compute and data, not headcount.

Leverage over headcount

Output scales with compute, not hires.

A small core sets direction and guardrails. The operator does the work, so output scales with compute and data, not headcount. The leverage is the loop, not the org chart.

The bet
Make autonomous media buying trustworthy, and available to every operator.
The operator
It ranks offers, allocates budget, refreshes creative, and acts.
Your accounts
Customers bring the ad accounts and the budget. Kerdixo brings the operator.
Public proof
A public live dashboard is the trust layer, not a marketing afterthought.
Autonomy
Shadow to Full Auto, always inside guardrails the customer approves.
Trust to demand
Proof in public turns transparency into trust, and trust into demand.
The data flywheel

It learns through operating loops.

Kerdixo improves through operating loops. Act inside guardrails, measure, learn, compound into the next decision. More operation creates more data, the data sharpens the operator, and a sharper operator earns more operation. The advantage cannot be bought, only accumulated.

Act Measure Learn Compound DATA FLYWHEEL

A few angles worth a story.

Threads we are happy to go on record about, each one on the record.

Radical transparency

The AI that publishes its own report card.

An operator that commits to showing its learning rate, guardrail saves, and incidents in public — in an industry where the norm is a black box.

Autonomy & trust

Would you let an agent spend your ad budget?

Kerdixo runs in Shadow until you trust it, then Full Auto inside guardrails it cannot cross. The trust dial, made explicit and reversible.

Agent-first company

It scales on compute, not headcount.

The operator is the workforce. A small team sets direction and guardrails while the system runs thousands of campaigns.

The moat

You cannot rebuild this in a weekend.

Anyone can wire an agent to an ad API. The moat compounds from day one: every decision the operator logs becomes an audited track record that cannot be bought or backfilled. The advantage is accumulated, not coded.

The agent economy

When agents start buying media from agents.

Brands are beginning to integrate through agents. The networks that win will prove where traffic came from. Kerdixo is built for that world.

The next category leaders will not be the most automated. They will be the most accountable. We are betting the company on proving it in public.Antoine Moreau, co-founder and CEO

Releases and coverage.

Kerdixo is in private beta and starting to tell its story. Announcements and media coverage are collected here as they land.

Press releases

The first announcement is in preparation. Check back soon.

In the media

No coverage yet. Writing about autonomous media buying or the agent economy? We would love to talk.

[email protected]

Use our logo in your story.

Download the marks below, or grab the full kit with colors, typography, and boilerplate. Vector and PNG versions available on request.

Kerdixo
Logo · on darkSVG ↓
Kerdixo
Logo · on lightSVG ↓
Carbon#06140F
Petrol#0A1E18
Mint#7CFFB2
Paper#F4EFE4
Download press kit (.zip)Spectral · IBM Plex Mono · Inter
About Kerdixo · boilerplate

Copy, paste, publish.

Kerdixo is the AI-native operator for affiliate media buying. Founded in 2026 and headquartered in the United States, it runs the work around the campaign that ad platforms cannot: ranking offers, allocating budget across a portfolio, supplying compliant creative, and calling lifecycle decisions, all inside approved guardrails. Every material decision produces a receipt, building a public operating record that makes verifiable autonomy the company's core trust infrastructure. Kerdixo is currently in private beta on Meta and TikTok.

Logos, product screenshots, and a one-page fact sheet are available on request.

Press contact

Talk to us.

For interviews, a product walkthrough, or background on the operator thesis.