The Flek Machine provides easy to use yet powerful toolset for AI Analytics.
Evolution of Analytics
Looking at the evolution of analytics as a whole, we see a drive towards intelligent processing that is necessitated by data complexity. In recent times, AI has taken inroads into the field and has made major contributions such as machine learning based analytics.
Problem of Machine Learning
Today, machine learning is difficult to apply within an enterprise because it is based on a paradigm that is uncommon to architects, data engineers, software developers and analysts alike. Instead of the streamlined workflow, they must now adapt to experiment-driven pipelines which do not fit with their database centered analytics.
For example, they can no longer design their models, upload data then "interact" by means of querying or mining to gain insight. Moreover, they need a diverse tools and techniques to prepare data, experiment with different algorithms, train the mathematical models, deploy for production and then run inference on new unseen data. All this makes ML difficult and very cumbersome.
Flek: A Platform for AI Analytics
To solve this problem, Flek offers a unified platform consisting of a running engine, server and programming toolkit that makes it easy to design models and then use them for exploratory and predictive analytics using the familiar database-like paradigm.
Beginning, users upload their data and then automatically build their probabilistic model using FlekML (the core engine). Following, they start "interacting" with their model by querying certain probabilities, mining interesting associations and rules, running complex probabilistic computations or by making various predictions. Thanks to the scalable architecture, users can also test their models locally, then deploy them remotely on the cloud similar to the way database models are deployed and shared today.