Discover the future of classification with AI solutions that excel on CPUs.

Lumina AI Awards & Coverage


Random Contrast Learning (RCL) is a new form of AI that ouperforms competitors with far lower costs in data, time, energy and hardware.

RCL™ accomplishes this feat without relying on resource-intensive neural network approaches.

Neural networks are designed to mimic the intricate structures of the brain. In contrast, RCL is designed to mimic the structures of consciousness – or more simply, how a person thinks¹. More specifically, RCL mimics how the mind draws distinctions and classifies its objects.  

Classification lies at the heart of AI and machine learning. It is the categorization of data, e,g., classifying biopsy images as malignant or benign. Although classification does not exhaust all AI tasks, all AI applications employ classification. 

Enter RCL, which is a universally applicable classification algorithm.

¹ This account of consciousness used in the development of RCL comes from the tradition of phenomenology founded by the philosopher and mathematician Edmund Husserl (1859-1938).

RCL can meaningfully expand the AI market and benefit organizations in several key ways:

Reduce Infrastructure and Energy Costs.

Most AI and neural-network based models run on GPU chips, which are very expensive, and energy-demanding compared to CPU-based computers. RCL can improve a wide variety of existing AI workflows by replacing the neural networks employed for classification. Use of CPU instead of GPU will greatly reduce an organization’s compute costs and carbon footprint.  

Increase both Speed and Accuracy.

RCL outpaces neural networks by orders of magnitude in data preparation, training, and inference speeds while utilizing widely available CPU hardware. 

Combine Models for Greater Efficiencies. 

Because RCL models can be combined without retraining, users can combine their models of the same data type and training parameters to derive insights without the risk of sharing their data. 

This will have broad applications and appeal across sectors, such as medical research. 

Increase Accessibility while Protecting Data Privacy.

In contrast to GPU, CPU is widely available. RCL’s CPU-based training makes high-quality AI tools accessible for a wider audience of individuals and organizations. 

Because RCL is optimized for CPU, a user can train their models on a local device, without sharing their data and without cloud access. For example, a farmer can analyze a vast set of farm sensor data at the “edge” on his or her home or business computer with no need to upload the data to the cloud.

Our products

Introducing PrismRCL™

Bring the power of Random Contrast Learning to your CPU-based Windows machines with PrismRCL™. With all of the benefits of RCL,™ utilize a simple command line interface to train, iterate and run inference against your models from your own Windows hardware.

Try it free for 30-days  and experience the future of machine learning for CPU.

Introducing the RCL™ API

RCLC is a version of Random Contrast Learning built for binary classification of PNG images, tabular and text data.

Utilize the RCL™ API to upload data, train your models, and combine disparately trained models in a single inference.

Important PrismRCL Update: Model Compatibility and Dual Version Support - Retain your existing models and explore 2.4.x features! Learn more about your options.