Gregor Kappler carries out psychometric research and data science consulting, and is founder of FilingForest, a julia-focused startup developing solutions for fast unbiased measurement in graph data.
Gregor was initially trained as a mathematician and psychologist, has implemented solutions for semantic text analytics for his PhD in 2007, and developed psychometric models for measuring with texts.
He has worked as a lecturer and researcher at the University of Vienna and the University of Jena and worked on a series of predictive analytic projects for software vendors and customers.
Gregor has switched to Julia from R in 2018, and is creator of the CombinedParsers package which provides parser combinators for fast, recursive and type-save parsing in pure Julia.
Parsers are programs to break apart strings that match a grammar
in order to transform into structured representations.
I demonstrate composing regular expressions, EBNF grammars and CombinedParsers.jl constructors to build a slick message broker system inspired by Apache Kafka:
Log lines and other messages are parsed to julia types.
Parsed instances are brokered by julia's multiple dispatch into different data sinks (git managed CSV, SearchLight.jl).