Scailable provides a unique platform to distribute compute-tasks over available compute-nodes. This allows us to carry out many modern Machine Learning and AI tasks (and pretty much anything you would typically carry out in a serverless environment) extremely efficiently and cheaply. To do so, we rely on WebAssembly; "a safe, portable, low-level code format designed for efficient execution and compact representation."
We focus on bringing machine learning to production using our platform. We can help you from the start:
https://sclbl.net/run/:cfidand you specify the requirements for your task.
Refer to our Frequently asked questions for more information.
Cutting-edge technology does not exist in a vacuum. We rely on many open source projects to make our stack work, and we collaborate with academic institutions and research labs. We benefit from collaborations with the Jheronimus Academy of Data Science,(itself a collaboration between the Tilburg University and the Technical University of Eindhoven) and the nth-iteration lab, a research lab developing machine learning and AI methods for personalization.
Our scailable-tasks are optimized and stand-alone. This allows us to distribute scailable-tasks to any available compute-node that is able to carry out the task. Our platform thus automatically, and in real-time, picks the compute node that is the most affordable. This can be a server at AWS, Google, or Microsoft, but it can also be one of our thousands of available web-nodes. Effectively, we provide serverless AI & ML on the edge.
Our approach liberates your compute tasks. They are no longer tied to the executing infrastructure, so you are not bound to a single cloud compute provider. Our unique model allows us to distribute your compute task to the most affordable compute supplier at that instant.
Our WASM optimization has increased the computational speed of evaluating Bayesian Additive Regression Models (see here for more information) by more than a 10-fold. Smartly distributing this optimized Machine Learning model over available compute-notes has cut the costs of running these models in production by over 60%.
Contact us if you want to get started right away.
Scailable-tasks are available to you at any time, at any scale, by simply POST-ing the details of your task to
:cfidspecifies the identifier of your compute task.
You are in control of the REST endpoints that enable your compute tasks. They are yours, and you can control access to them. You can use them solely for your own purposes, or you can open them up for others. This even allows you to monetize your tasks by allowing third parties to use them.
We have a number of publicly available task can be consumed. For example:
|Tweet emotion detection||454sday-511a||Text / plain|
|Optimal route selection||99nfh-33nz||JSON|
We work together with one of the Netherland's largest telecom operators to predict churn at scale.
We are working towards different solutions to put insurance fraud detection into production.
Our optimized machine learning models allow easy and secure classification of medical images.
Our platform provides compute suppliers with a unique model to monetize their compute power. Any laptop, phone, supercomputer, or server can contribute to executing compute tasks. Obviously, as a supplier, you can earn by supplying your compute resources.
Our distributed model provides web content providers with a novel mechanism for monetizing their content. Adding a simple
.js snippet to your website turns the machine of any visitor to your website into a compute node; a resource you as a content provider can monetize. Contact us if you want to join!
If you want us to help you turn your Machine Learning or AI task into a Scailable task and allow you to put it into production at scale, do let us know. Fill out the form below, or give us a call at (+31) 0619196802.