Flek makes data science and AI Analytics easier by bridging the gap between data and insight. It carries out this by:

• Helping organizations leverage their existing expertise (people), data (SQL databases) and workflows (pipelines).

• Enable AI citizens like data scientists, analysts, programmers and statisticians to run their queries, mining and prediction tasks that use the same share central models - similar to how SQL databases are used today.

• Simplifying machine learning and model maintenance using auto-learning and self-adjusting models running on a central probabilistic engine.

• Integrating in one Python Toolkit: data preparation, model building, querying, mining, auto-discovery , visual exploration, prediction and recommendation.

• Enabling data science citizens to interpret and explain the results of their analytical activities during exploration, discovery or prediction.

​

Examples

Let's suppose in a large organization an analysts wants to explore customers' profiles; while the loan manager wants to categories new loan applicants based on their riskiness; and maybe the product manager wants to predict which customers might be interested in a new line of credit.

 

​Flek makes it simple to conduct these diverse tasks in a standardize and holistic approach. The analysts can dive into details of customer profiles to uncover probabilistic patterns or spot anomalies. While the loan manager can get more insight by running a classification task using the probabilistic model (running on central engine) to segment loan applications. Whereas the product manager can run a prediction activity that takes existing customers as input and then predicts the level of interest in the new line of credit (given their credit history and profiles); a kind of simulated recommendation use case.​