Reinforcement Learning MLOps Platform



1. Objective

AgileSoDA has a Reinforcement Learning based MLops Platform Product. Creating an MLOps platform designed specifically for reinforcement learning, making it easier to manage simulators, environments, and numerous hyperparameters.

2. Role

The main skills I used in this product are : MLOps, PyTorch, NNI, Linux, RL, Customer Support.

  • Unit Test : Conducted unit tests, a testing method used to verify the functionality of specific components or modules within the platform.
  • Development : Development of test codes for reinforcement learning algorithm libraries, version management of libraries, writing Dockerfiles, Git Flow collaboration. Development of Python Web Framework based simulator
  • Manual : Created product manual and educational materials.
  • Customer Support : Provided customer support for education purposes and developed examples cases for customers' easy use of the product.
  • 3. Supplementary Materials

  • More information about the product can be found here