View on GitHub

---

Applied ML

You are here: vertex-ai-mlops/Applied ML/readme.md

Applying models to enhance workflows is Applied ML – often referred to as AI.

Applied ML

In some cases, a workflow requires model output, which might seem related to the serving topic covered in MLOps/Serving. However, it's actually the next step: what you do with the response. For instance, in forecasting, you might use the forecasted horizon to change how you interact with your customers or your supply chain. A great example of applied ML is generative AI, where a very general model serves many purposes and can be coerced into conducting actions, such as agents. Another area of applied ML involves combining or specifically applying techniques to achieve an objective. This is the purpose of this area of the repository.

The work in this part of the repository will make use of:

• Multiple Frameworks which are covered directly in Framework Workflows
• MLOps techniques which are covered in detail in MLOps

Application Areas:

• Forecasting
• Generative AI (GenAI)
• AI Agents
• Anomaly Detection
• Optimization

Solution Prototypes

Some workflows require complex applications or the combination of multiple applications. These solutions can solve complex business scenarios.

• Solution Prototypes