Building Scalable Prediction Services in R
0 likes10,412 views
The document discusses scalable prediction services for real-time fraud detection using R models, focusing on challenges like the lack of DevOps compatibility and limited model types. It proposes solutions such as embedding R models within middleware to enhance management and deployment processes while leveraging existing DevOps strategies. The conclusions highlight the rapid deployment of R models in scalable environments, effective integration with data science workflows, and potential future serialization of models for compliance.
Building Scalable Prediction Services in R
- 1. Scalable Prediction Services with R
- 2. #RstatsNYC @Socure • Real-time fraud detection service using social and online data. • Predictive R models. • Latency SLA with customers. • Model versioning. • Zero-downtime updates.
- 3. #RstatsNYC @Socure Challenges • R not dev-ops friendly. • Enterprise prediction services a large commitment. • Enterprise prediction services offer limited model types. • Transferability and transparency of models. • Vendor lock-in.
- 4. #RstatsNYC @Socure Solution • Embed R models within dev-op friendly middleware. • Management, deployment, integration leverages existing dev-op processes. • Service scaling using established strategies and methods.
- 6. #RstatsNYC @Socure <model> name = generic version = 20150215 <file> gen_20150215.rds readRDS() saveRDS()
- 7. #RstatsNYC @Socure Rook http://…./model/20150215 <model> name = generic version = 20150215 name version model 20150215 Model Map <file> gen_20150215.rds readRDS() saveRDS()
- 8. #RstatsNYC @Socure Rook http://…./model/20150215 <model> name = generic version = 20150215 name version model 20150215 Model Map predict() <file> gen_20150215.rds readRDS() saveRDS() JSON
- 9. #RstatsNYC @Socure POST generic/20150215 Rook Rook Rook Rook fork() ……..
- 12. #RstatsNYC @Socure Virtual Machine Docker Public Repository ECS ElasticBeanStalk R R RR R R
- 17. #RstatsNYC @Socure Conclusions • Rapid deployment of R models in a scalable robust environment. • Directly leverage R models developed by data scientists and analysts. • Apply existing dev-ops processes for testing, monitoring, scaling, alerting of predictive models. • Possible use of PMML to serialize models in future for compliance.
- 19. #RstatsNYC @Socure We’re Hiring http://www.socure.com/hiring Director of Data Science Senior Data Scientist Director of Engineering