Skip to content

Marketplace

Open your fleet to research. Capture revenue you would otherwise miss.

Allow Embeint research partners to work on your production devices. When they build something your customers will pay for, you choose whether to offer it, and you are who the customer buys from.

  • You choose the apps
  • ·The customer buys from you
  • ·You keep 20% of app revenue

What you actually do

Embeint works with researchers to foster R&D on your production devices. You allow some of your fleet to be used for that work. They innovate on hardware you already have. You commit to a path to sell apps to your customers. You give up nothing you need to fear.

You still choose

You choose which apps you offer. The livestock producer, or whoever your customer is, buys from you, not from an Embeint app store.

How value splits

You keep 20% of all Marketplace app revenue on your device fleet. The researcher is paid for the intelligence. Embeint takes a smaller platform fee. You did not fund the R&D.

What an app is

On-device intelligence that matters.

Typically tinyML or AI on the device that alerts the customer to an issue of material importance. Example: Precision Livestock on CERES TAG detecting a disease or health condition. The producer buys from CERES TAG.

Live path

CERES TAG is the working example.

Researcher Toolkits collect data on production hardware. If customers will pay, the device partner can sell the capability. Coming on the same pattern: asset tracking, and companion-animal health, without naming unannounced OEMs.

Marketplace economics

Researchers innovate on your fleet.

Embeint works with researchers to foster R&D on your production devices. That work runs on hardware you already have in the field. When it becomes an app your customers will pay for, you choose whether to sell it. You keep 20% of all Marketplace app revenue. You did not fund the R&D.

Illustrative example on a 250,000-device fleet. Customers take different apps. Some take more than one.

  1. Step 1

    A catalogue, sold by you

    Embeint works with researchers on your hardware. You choose which apps to list. The customer buys from you, under your brand, not from an Embeint store. One device can run more than one app.

  2. Step 2

    Different mix per customer

    • Disease / health90,000 · $360,000
    • Behaviour140,000 · $280,000
    • Heat detection80,000 · $240,000
    • High-rate GPS50,000 · $75,000
    • Welfare160,000 · $160,000

    Not every device takes every app. Take-up differs. Devices that run two or three apps earn on each one.

  3. Step 3

    Infuse meters each app

    • Counts devices that ran each app
    • Bills through your customer relationship
    • Adds the stack, then splits it
  4. Step 4

    You keep 20% of the stack

    $223,000

    A new app on the catalogue is a new line. It does not take a new R&D team of yours.

And again next year, for every app still running on a device.

$1,115,000 app revenue

on a 250,000-device fleet. Five apps, mixed take-up.

20% yours

on every app that runs

$223,000 is yours

illustrative, recurring while apps stay enabled

Embeint brings the researchers. You open the fleet. You did not fund those algorithms. More apps on the same devices compounds revenue without a second R&D budget.

This example uses a 250,000-device fleet. App names, rates, and take-up are illustrative. Your 20% share applies to all Marketplace app revenue on your fleet. Talk to us to open a fleet.

If you have intelligence

Start at Researcher Toolkits.

Most university research never ships. Here the work is already on production hardware. We match the right device before anything is listed. Owners of tinyML from other ecosystems can still talk to us about porting onto partner devices.

Go to Researcher Toolkits

If you run devices

Your customers get new intelligence.

Embeint works with researchers to innovate on your fleet. You did not have to fund the R&D. Capital stays on growth. A new algorithm is a new line for the researcher, and a new offer you can choose to sell.

Talk about opening a fleet

“An algorithm that never reaches a device is just a demo.”

Robert Sandona · Senior Cloud Engineer

Open a fleet, or list an algorithm.

Device partners talk to us. Algorithm creators start at /research.