High Performance Data Mart using Event Plots

Reporting data for finite state machine transactions like calls require exhaustive data mining into 10s and 100s of parameters for 1000s of calls per hour. These postmortems require heavy processing and can demand millions of maps per reduction. The presentation is a case study and demonstration of near-realtime asynchronous event processing based plotting of events and states of FSMs. Thus creating semi-cooked event plots delivering lightening fast reports with extensive customizability.

Slides
Video

About Chaitanya

I am an Erlang Programmer and solution architect primarily for the telecom domain. I have 14 years of experience in Erlang programming. Have been designing high performance low latency telecom grade solutions for 13 years. I have designed and developed roaming solutions, prepaid billing systems, high performance middleware and cloud based solutions. These days I am working on cloud based call center solution. As an architect my primary choice of language for the backend and middleware has always been Erlang.

Twitter: @chaitanyach

Back to conference page