OPEN COGNITIVE FRAMEWORK
A NEW APPROACH
TO LEARNING @ SCALE
It's not the parts,
it's how they are brought together.
WHAT IS A COGNITIVE FRAMEWORK
We get it. It sounds "buzzy" - but it's not.
A "cognitive framework" is a newer approach for how learning is designed and deployed.
In essence, we design our systems to "think" like we do - working backwards from business goals, to the insights required to drive them, the models required for insights, and ultimately the data/context that fuels it all.
THINKS LIKE WE DO
It implies alot - because we know alot.
Our customer communities can change day by day. New products are introduced while others go NLA.
Everything that makes retail...well...retail is factored into the Cognitive Framework - as it's the only way to make learning continuous.
"Features" are the data that fuels models.
In traditional learning, models are changed to support changes in the data and feature - a manual process that drives costs.
In a Cognitive Framework features are crafted from data. Meaning the features are designed to allow the model to factor in changes.
To achieve this, we created data contextualization. The process in which the meaning of data
A key notion in creating a Cognitive Framework is having a defined "big picture." What we mean by this is have an understanding of scale.
A simple example is churn. Churn to some means across your retail brand, to others it's a store or website event, to other it is within product lines. To us, it's all of the above.
Because of that - "transactional" churn risk starts with product line and describes the customer relationship upwards - through the sales channels and upward to the brand.
Big picture is going granular.
OUR COGNITIVE FRAMEWORK
It's worth a reminder - our Cognitive Framework is part of OpenINSIGHTS.
Our cognitive framework methodology is included in OpenINSIGHTS and truly is reflected in all that we do.
From data onboarding, data contextualization through our modeling (even our neural network) - Early Access clients not only get the infrastructure, but their data teams get trained on how to extend it.