Rulex AI

At the core of Rulex is a family of proprietary machine learning algorithms which are different from conventional machine learning algorithms, like Decision Trees, Neural Nets, and others.

Decision Trees and Random Forests produce predictive models that are readable, but not understandable.

Rulex allows overlapping rules for simpler, more understandable predictions than Decision Trees.


Neural Nets, as well as Support Vector Machines and most other popular math-based algorithms produce math functions that are neither readable nor understandable.

Unlike the binary trees and solving functions of conventional algorithms, Rulex works the way the human brain works, using learned logic to make decisions.

IF (customer_province in {A, B, C, D} AND damage_class in {1} AND Number of days between policy start and date of accident <= 371) THEN Fraud = Yes

IF (customer_province in {E, B, C, F} AND Customer age > 48 AND Number of days between date of accident and complaint  > 1) THEN Fraud = Yes

IF (customer_province in {G, H, I, J, K, L, M, N, B, O, P, Q, R, S}) THEN Fraud = No

IF (Number of days between date of accident and policy end <=  2) THEN Fraud = No

For migrating to Rulex and supporting legacy machine learning workloads, Rulex also supports most conventional machine learning techniques with a suite of popular algorithms with proprietary enhancements for higher performance and greater ease of use.

Machine learning for AI has never been easier, more economical, or more effective than it is with Rulex.  Conventional methods are iterative, complex, and not transparent.  Rulex is automatic, simple, and clear.

Neural Networks, SVMs, etc.

Users are data scientists

Requires math and programming

Manual data exploration

Manual, experimental modeling

Math function-based models

Black box predictive decisions

Prediction on central servers

New predictions require recoding

Rulex Logic Learning

Users are business/process experts

No new skills required

Automatic data discovery

Automatic model building

If-then rules-based models

Self-explanatory predictions

Prediction on edge devices

Update predictions w/o recoding

Rulex for Manual and Automated Machine Learning