Frank Font- FINRA, ,Yan Cheng-FINRA, David Llanos-FINRA
Title: Deploying R Models to Implement REST Production Endpoints
Title: Deploying R Models to Implement REST Production Endpoints
Abstract:
Abstract:
The R suite of tools is the go-to choice for many ML/statistical modeling professionals. Sometimes models built using R live only through the prototype stage, converted into another technology for rollout into production. We wanted to explore the possibility of using R as more than just a prototyping tool, we wanted to see if we could deploy and use it in production as the workhorse for the ML/statistical models needed by a new application our team was building. We found success in our team through the use of plumber to turn the R models into REST endpoints and an infrastructure/JSON configuration scheme to enable rapid
The R suite of tools is the go-to choice for many ML/statistical modeling professionals. Sometimes models built using R live only through the prototype stage, converted into another technology for rollout into production. We wanted to explore the possibility of using R as more than just a prototyping tool, we wanted to see if we could deploy and use it in production as the workhorse for the ML/statistical models needed by a new application our team was building. We found success in our team through the use of plumber to turn the R models into REST endpoints and an infrastructure/JSON configuration scheme to enable rapid
deployment of version-controlled models across environments.
deployment of version-controlled models across environments.
Team of Yan Cheng, David Llanos, and Frank Font will share their process.
Team of Yan Cheng, David Llanos, and Frank Font will share their process.