Lakes, clouds and unicorns: A tale of data science
Rank Group’s head of data innovation, Jan Teichmann, on how the industry should be leveraging its data
Many operators breathe a sigh of relief on completing their digital transformation, and slowly recover from the challenges the change brings. But there is no time to take a breather as the next major disruption is upon the gaming industry: the data transformation.
It is almost impossible to miss how the data field gained some new buzzwords. It used to be ‘big data’ but, as of late, it is dominated by data lakes, machine learning, AI, blockchains and data science unicorns. Is this just the latest hype or is the tech landscape undergoing a major revolution and is data finally taking over?
Gaming is a data-rich industry and has always been, but the utilisation of that data has traditionally been poor, with challenges arising from the great number of third-party legacy systems and their limited integrations. With the desire to improve the effective use of data and a de facto requirement for predictive responsible gambling models nowadays, operators often have dedicated data science or advanced analytics teams. In a moment of honest self-reflection, you might want to ask: “Are these teams delivering to their full potential and turning the dial for us?” You are not alone if you feel a little disappointed after such an assessment.
The pioneers in the data field – with entire armies of engineers – have successfully proven the transformative impact data has on business and customer experience, and yet also highlighted how disruptive the data transformation is on the technology currently powering most data infrastructure. The fact is that the data transformation is, at its core, a tech revolution, and most operators do not have the resources or skills to meet the requirements of their data science teams to bring models into production, and certainly not at scale.
The gaming industry has been late to the digital transformation and is facing some of the bigger challenges when it comes to their imminent data transformation. But that is no excuse for delay as, in all honesty, the industry’s typical data infrastructure is already feeling the strain.
A strong future
The technology exists today to run a modern data infrastructure to support data science in production at scale, and it is now also accessible to the gaming industry from the wide range of PaaS offerings. The foundation of a future gaming business is a robust data infrastructure and the key to scaling data science will be the flexible rendezvous architecture. In a successful data science environment, the boundaries of development and production are increasingly blurry.
A single business objective will see many different incumbent and challenger models, and a requirement to retrain these models in a constantly changing environment with their continuous benchmarking, monitoring and production hand-over. These are requirements giving any IT department a good run for its money.
At the Rank Group we are revolutionising our data infrastructure with just a small agile team, building a modern data lake and flexible rendezvous architecture for data science. The challenges we are addressing are the increasing heterogeneity of the data science toolkits, the required elasticity and flexibility of the data infrastructure, and the end-to-end automation of the complex data science model life cycle. The secret to the successful delivery of that project is the use of PaaS, properly decoupled micro services and the valuable know-how which exist within the open source data science community.
There are no good reasons for the gaming industry not to face the new data reality head on for the benefit of the industry and the customer. Don’t delay, join the Rank Group on the other side of the rainbow where data flows and unicorns roam the land of data science.
Jan holds a PhD in Mathematics from City University London and offers a strong background in machine learning, statistical modelling and programming. He uses his skills now at the Rank Group to build a production environment for Data Science and to drive personalisation of their digital gaming experience.
