ElixirConf 2023 - Sean Moriarity - MLOps in Elixir: Simplifying traditional MLOps with Elixir

Discover how Elixir simplifies traditional MLOps with NX serving, distributed abstractions, Livebook, and telemetry events. Learn how Elixir's features empower developers to build robust machine learning applications.

Key takeaways
  • Elixir simplifies aspects of machine learning deployments, especially inference and serving.
  • NX serving is a powerful abstraction for deploying machine learning models.
  • NX serving handles dynamic batching, concurrency, and load balancing for you.
  • NX serving can be used with any numerical definition, including models trained with Axon, Scholar, XGBoost, and ONNX.
  • Elixir’s distributed abstractions allow for distributed pre- and post-processing of data.
  • Elixir’s Livebook can be used to connect to production applications and explore data in real time.
  • Elixir’s telemetry events can be used to debug and monitor NX servings.
  • Elixir’s small but powerful features allow developers to punch above their weight when building machine learning applications.