R&D Challenges

In any unknown/not invented technological process there is an element of discovery and quantification.

We’ve been working closely with our partners at Totem Learning to identify how we can build a robust, scaleable , efficient and effective mechanism for their future vision of learning and recruitment.

Generalised Data Management (at scale)

Building a system that can handle extreme ingress rates while allowing data mining and extrapolation of data points is typically handled by generalising data. 

Good examples of this are our architectural designs/implementations for massively concurrent online videogames where 100,000 players enter and leave the system per second, or our advertising intelligence work that determines the best messaging for a viewer dependent on their recent behaviours/traits by calculating segments in realtime.

Individual’s Data Management (at scale)

Building a system that tracks and traces individual records, compares them to cohorts to allow segmentation (both steered and fully automated) has a requirement for managing each individual’s record separately.

Good examples of this are tools such as NSE.gg or the esport tournament systems we’ve built: each allowing individual’s progress and traits to be managed, viewed and collated in batches (largely for cost efficiency/resource management reasons).

What’s Special about Totem’s requirements?

Totem Learning’s system requires both the aspects of a generalised system and the individual record management.  This presents new challenges because the best parts are required from each architecture without the downfalls/caveats that each normally presents.  This makes this research (and hopefully the product that will spring from it) both innovative and a hot bed of questions that can only be answered through experimentation and careful deliberation of the data and flows.

Code Wizards’ background in building massively scaleable platforms gives us the right baseline to attack this problem.  From the analysis we can work out the correct building blocks, where existing software packages/architectures are viable and where new technologies need to be invented to fill those gaps.