Presentation of development flow dedicated to Machine Learning internal and external projects including datasets storage and versioning via Quilt (as an alternative to DVC), experiments scheduling and infrastructure maintenance via Polyaxon, continuous Integration using CircleCI and Configuration & deployment using Docker.
The talk is about presenting a development flow dedicated to Machine Learning internal and external projects, practiced in our company (Netguru). The flow includes following parts:
Presentation describes of all components and explains how to put it together. Given Machine Learning flow impacted on on our projects' success stories. Further part of the talk would explain why it was so crucial to build and maintain it.