Phoenix Arize
Phoenix Arize is a platform focused on AI observability and monitoring. It helps organizations monitor, manage, and improve the performance of machine learning (ML) models in real-time. The goal of Phoenix Arize is to provide tools for tracking the behavior of AI models once they are deployed, ensuring they continue to operate optimally, and detecting potential issues such as model drift, bias, or degradation.
Key features of Phoenix Arize include:
- Model Performance Monitoring: Continuously tracking the accuracy and performance metrics of deployed models.
- Drift Detection: Identifying when the behavior of the model changes due to new data or shifts in underlying patterns (often referred to as concept drift).
- Bias Detection: Checking for unfair biases that may arise in AI predictions.
- Data Observability: Offering insights into the data that is being fed into AI models to ensure consistency and quality.
- Actionable Alerts and Insights: Notifying teams when certain thresholds are met or anomalies are detected, allowing for faster interventions.
Configuration¶
The below content provides a sample notebook which can be used to create and show tracing and evaluation data within the Phoenix app (i.e. their SaaS offering).
The following accounts are needed for this exercise:
Within you Rafay Kubeflow based MLOps environment:
- Navigate to Notebooks
- Click New Notebook
- Enter a name for the notebook
- Select JupyterLab
- Set the minimum CPU to 1
- Set the minimum memory to 1
- Click Launch
It will take 1-2 minutes to create the notebook.
- Navigate to Notebooks
- Click Connect on the previously created notebook
- In the left hand folder tree, click on the upload files icon
- Upload the previously downloaded phoenix_arize.ipynb file
- Double click the phoenix_arize.ipynb file in the folder tree to open the notebook
Credentials¶
You need to provide credentials for the notebook to connect to your Phoenix and OpenAI accounts.
- Update the section in the notebook that contains ENTER PHOENIX API KEY with the value of your Phoenix API key found here and then selecting Keys from left hand tree
- Update the sections in the notebook that contains ENTER OPENAI API KEY with the value of your OpenAI API key found here
- Click the Restart kernel and run all cells icon
Model Tracing & Evaluation¶
The tracing and evaluation data will be sent to your Phoenix account and can be accessed by going to the default project