| Overview: |
As games grow in complexity and size, the systems they rely on become increasingly difficult to masterand so does their technical optimization. Often handled by experts at the end of a production, the analysis and prioritization of such optimization tasks is a long and tedious job.To overcome this problem, we developed an automated solution leveraging big data platform. We collect, store, transform, aggregate and make profiling data available to developers through standard analytical dashboards and queryable databases. We also engineered machine learning models to prioritize which system would be the best to optimize.We were able to reduce the time spent on analysis and prioritization from hours to minutes. Moreover, profiling questions are now more accessible thus it empowers all developers on their system's performance, making profiling concerns not only an expert field.
|
