What the Flek is the Flek Machine?

At the core of GoFlek’s AI software lies the revolutionary Flek Machine, a unique probability engine that stands as the first-ever built Probability Machine. This groundbreaking technology is not merely an incremental improvement to modern machine learning but a transformative innovation for the future.

Unique features

Transparent
The Flek Machine is not a black-box: all results can be demonstrated, traced-back and explained.

Deterministic
Given a certain combination of inputs, the Flek Machine will always provide the same outputs and results.

Versatile
The Flek Machine integrates several modules to perform various tasks (insight discovery, predictions, recommendations…)

Sovereign
GoFlek has been built in Canada, owns 100% of its code, and does not have any third-party dependencies.

Sustainable
GoFlek has been optimized at the very heart of its algorithms, for maximum power efficiency.

The Flek Machine integrates 3 key components: FlekML Engine, Server, and the Python Toolkit. Together, they form a unified framework that allows businesses to auto-build probabilistic models on the fly, and conduct both exploratory and predictive analytics, all within one cohesive tool.

  • Flek ML Engine

    The FlekML is the core machine learning engine with a core Probability Machine inside. It builds models by learning Nuggets from the semi-structured data. It then stores these models into memory, files or a database. FlekML also serves user the stored Nuggets when they query and mine or when they search for probabilistic patterns.

  • Python Toolkit

    The Python Toolkit and library for AI analytics includes a suite of components that enable users to: query, mine, discover, explore, make predictions or run recommendations. It also includes APIs to help prepare data, visualize and interact with the stored models as well as components to validate and trace the results of a prediction task.

  • Flek Server

    The Flek Server runs the core engine in multiple modes: standalone, local or remote (cloud), which makes it easy to test locally or deploy on a larger scale on the cloud. To safeguard its data, all communication is conducted over HTTPS with built-in security safeguards.