BaynoteONE RTX
Real-time Experience Engine

The BaynoteONE Real-time Experience Engine (RTX) is a modern machine learning engine designed to put big data to work in real-time. More than just big data aggregation, the purpose of RTX is to make big data insights actionable.

A New Big Data Modeling Approach

Battle Tested Algorithms – The BaynoteONE RTX engine and algorithms are battle tested and proven to drive value. Some systems pick from a number of simple algorithms which can produce poor results on long-tail items and create an unpredictable customer experience. BaynoteONE uses a unique composite algorithm framework that leverages the best of item context, site behavior, and individual customer behavior and customer preferences to drive relevant, personalized results. Baynote holds a number of unique algorithm patents and applied for two new patents in 2014.

True Machine Learning – Most predictive analytics systems simply exploit historical trends found in the existing data, they don’t test and explore to see how customers will react. The BaynoteONE RTX modeling engine uses reinforcement learning techniques to actively test and explore to find new, meaningful results more quickly that continue to improve engagement and conversion to drive more revenue.

Online & Offline Data Integration

Real-time data – BaynoteONE uses a unique data tag to collect behavioral and customer data from the real-time experience. This data includes the items or pages that a customer is most interested in, purchases, user entered data, search terms, and customer attributes available as part of session data.

Periodic data feeds – BaynoteONE also takes data into Hadoop via periodic data feeds using Pentaho ETL. Periodic feeds include product and content catalogs, CRM profiles, and offline purchases from stores or agents. Periodic feeds are integrated with real-time data to create a more comprehensive picture of customer behavior and profiles that enhance the modelling process.

Machine Insight in Real-time

Real-time Performance – The BaynoteONE RTX real-time interaction server is a unique BaynoteONE technology that puts machine-learning models into memory for real-time access in milliseconds.

Minimal Latency – “Over the wire” communication between SaaS technologies and client systems can create delays for page load time or email responses. BaynoteONE RTX real-time interaction server is a virtual server that can be hosted almost anywhere including by Baynote, on a cloud service like AWS or co-located with client servers. This flexible architecture allows BaynoteONE to scale to meet any client need with extremely low latency.