An example business use case and reference architecture
We will focus on a computer vision (CV) use case to identify objects in images. Extending the CV use case we implemented in Chapter 15, we will build our CV model via an MLOps pipeline that incorporates the majority of the main topics from the earlier chapters in this book, such as the following:
- Data preparation, cleaning, and transformation
- Model training, deployment, inference, and evaluation
The novel approach we are using here is to utilize generative AI to generate the data that will be used to test and evaluate our model. Before we start building, I’ll present additional context on why I chose to include this use case as an example.
Additional background on our use case
Throughout this book, one topic has come up more frequently than any other: data is often the most important factor in training high-quality ML models. Without adequate data, your ML project will not succeed. We’ve...
