Razorthink: Creating the First Artificial Intelligence Data Scientist
Jack Porter, President & CEO
The goal is to turn data into information, and information into insights. This adage by Carly Fiorina aptly communicates the significance of data. While it’s crucial for an organization to have access to data, it’s even more critical to have the ability to analyze that data to drive actionable insights. Today this is a major concern for businesses due to the exponential growth of Big Data, which has surpassed human data scientists’ capacity to analyze it. Many large-scale data analytics projects are stuck in pilots or are not feasible using human data scientists or machine learning.
In such a scenario, Artificial Intelligence is poised to be the new UI for the enterprise. “Companies need to rapidly analyze volumes of data in real-time and at scale—something not possible with traditional analytics systems and human data scientists,” says Jack Porter, President & CEO of Razorthink.
Established in 2011, Razorthink is a pioneer in Enterprise Superintelligence and has developed the world’s first AI Data Scientist, Big Brain. It solves two massive problems in Data Science. In today's large enterprises, core data systems produce terabytes of information every day. This is impossible for a human data scientist to comprehend, let alone identify meaningful patterns. In addition, to model this data they must use sophisticated model architectures, with often times hundreds of layers deep. To select and optimize these models is way beyond even the best mathematicians’ capabilities.
Big Brain excels at identifying complex problems with exponential variables over time. Unlike simple Machine Learning, Big Brain’s Deep Learning AIs can identify patterns not described by humans, non-linear patterns, and patterns with thousands of variables.
Razorthink has combined High Performance Computing (HPC) with Advanced Deep Learning Neural Networks to create Big Brain, the first AI Data Scientist
Porter adds, “Razorthink’s unique Artificial Intelligence techniques make it possible to build and continually optimize data models using machines, rather than having humans identify relevant features in raw data.” Razorthink Big Brain takes the load off of human data scientists by selecting, tuning, testing, and optimizing data models much faster, for much larger data sets and at levels impossible by humans. It allows businesses to harness the power of Big Data in ways that were previously not possible.
Specifically, the Big Brain Data Optimizer integrates, refines, and intelligently encodes the data. Data is then formulated into tensors (multi-dimensional arrays) for AI data modeling. Big Brain’s Model Optimizer selects the ideal model and then automatically tunes the parameters and hyper-parameters for each model through the Model Optimizer. Then Big Brain’s Deployment Optimizer tells users how to configure the system in terms of number of services, amount of storage, software needed based on the task it has been assigned. Finally, Big Brain deploys the model to the cluster using sophisticated data parallelization and model parallelization.
Razorthink works with some of the largest banks and telecommunications companies in the world. With the company’s unique Customer Expert AIs, clients can predict customer behavior in real-time, giving them the opportunity to change the outcome.
Razorthink Expert AIs provide predictions on customer behavior with higher accuracy beyond what is possible with human data scientists. “Industry reports estimate that banks are losing customers at an average rate of 12.5 percent per year, while average acquisition rates are at 13.5 percent,” says Porter. To address the situation, the company offers a Customer Expert AI for the Banking industry. Razorthink is also formulating Expert AI solutions for Telecommunications and Retail, which will reach customers this year.