Interview
summary
A Bayesian experimentation platform at Udemy handles Bayesian experimentation at scale by implementing bootstrapping with Spark and Scala using algorithmic optimizations to turn an otherwise expensive resampling method into a practical analysis pipeline, delivering measurable business-value metrics. Work addresses framework, analysis, and education challenges, favors functional programming and cats for clearer reasoning, and reports transfer of ideas from akka/streams-style patterns and contributor conversations into production code.
AI-written summary · about
chapters (5)
related talks
37mRealtime Bayesian A-B testing with Spark Streaming
04minterview
01h 30mBayBE: A Bayesian Back End for Experimental Planning in the Low-To-No-Data Regime
01h 03mExperimentation in Data Science
24mJoseph K. Bradley, Software Engineer, Databricks Inc. @ MLconf SEA
05mInterview
30mProbabilistic modelling with Scala
25mExploring Tradeoffs in Experiment Management as part of AI Platforms