Learn AI · Level 3 · Tech and tools · step 5 of 16 · word 1 of 2
What is MLOps (and LLMOps)?
MLOps is DevOps for machine learning, the practices and tools that get models into real use and keep them working, from training and testing to monitoring and retraining.
In plain English
Training a model is the start, not the end. MLOps borrows from DevOps: automate the steps, track every version of the data and the model, test before release, and watch the model once it’s live. Then retrain it when it slips, which ordinary software never needs.
An example
LLMOps is the same idea for chatbots and other apps built on large language models. Teams track versions of their prompts as well as models, run evals before each change, and watch cost, speed and guardrails once it’s live.
Watch out
In a well-known 2015 paper, Google engineers warned that machine learning systems carry large ongoing maintenance costs. Most of the work sits around the model, not in it. See every step in how a machine learning model is built.