Abstract: The advent of ever more affordable and powerful cloud computing has offered new opportunities to crunch massive amounts of data. But when your domain problem requires computation loads that scale exponentially, no amount of parallelization or cash will save you. In this talk we’ll explore indexing strategies to reduce exponential Big O problems to reasonable workloads through the polymorphism offered by the GiST index type in Postgresql and look at how to incorporate these strategies with other tools to build robust and performant analysis pipelines.
Bio: CTO and co-founder of dataPlor. Co-founded BrandFolder ($155M exit) and has 10+ years of engineering experience in the startup space, in addition to 5 years as a data analyst in financial litigation consulting.