ML Observability Workshop Series
Taking a model from research to production is hard — and keeping it there is even harder! As more machine learning models are deployed into production, it’s imperative to have the right skillset to monitor, troubleshoot, and explain model performance. That’s why Arize is hosting an ML Observability Workshop Series to help data scientists and ML practitioners gain confidence taking their models from research to production.
Each week, we will cover a key area of ML Observability and practical applications. You will gain a hands-on understanding of how to identify where a model is underperforming, troubleshoot model and data issues, and how to proactively mitigate future degradations.
Upon completion of this series, you will receive a ML Observability Fundamentals acknowledgement for your new skills!
Register For the Series
This series occurs every Tuesday at 9am PST/12pm EST
Arize sessions:
ML Observability 101
Do you feel confident when taking models from research to production? Do you want to learn how to root cause model failure modes in production?
May 24th, 2022
, 9:00am
-9:30am
PST
Read more →
ML Observability 101
May 24th, 2022
, 9:00am
-9:30am
PST
Join us to get setup with your free Arize account to start monitoring your ML models today! This hands-on workshop educates participants on ML Observability by demonstrating how teams across industries are using Arize to monitor, troubleshoot, and explain models in production. Surface drift, data quality, and data consistency issues in production, while learning how to resolve performance degradation using explainability and slice analysis. Afterwards participants will be able to leverage ML Observability in their own models using the Arize platform.
Performance Tracing with Arize
Do you know the exact moment your models in production start to degrade? Is your team equipped to quickly and correctly root cause and resolve performance issues in production?
May 31st
, 9:00am
-9:30am
PST
Read more →
Performance Tracing with Arize
May 31st
, 9:00am
-9:30am
PST
Join us to learn how to leverage Performance Tracing to identify when and where your model first starts underperforming. This hands-on workshop educates participants in performance troubleshooting workflows by demonstrating how teams across industries are using Arize’s state of the art Performance Tracing capabilities in their production workflows. Afterwards participants will be able to apply performance monitors and leverage Performance Tracing in their own models using the Arize platform.
Drift Detection with Arize
Do you know the exact moment your production models start to decay? Are you able to correctly identify the type of drift which occurred and resolve it immediately?
June 7th
, 9:00am
-9:30am
PST
Read more →
Drift Detection with Arize
June 7th
, 9:00am
-9:30am
PST
Learn how to identify the major causes of drift in your ML workflow. This hands-on workshop educates participants in the nature of drift and the forms it takes in production by demonstrating how to use drift detection to prevent performance degradation. Afterwards participants will be able to apply drift monitors into their own models in production using the Arize platform.
Data Quality Management with Arize
Are you correctly identifying data quality issues in production data? Do you know the exact moment when your model inferences start to act up?
June 14th
, 9:00am
-9:30am
PST
Read more →
Data Quality Management with Arize
June 14th
, 9:00am
-9:30am
PST
Learn how to identify the major causes of data quality issues in your ML workflow. This hands-on workshop educates participants on how to identify and resolve bad data inferences in production by demonstrating how teams across industries are using ML Observability. Afterwards participants will be able to apply data quality monitors into their own models in production using the Arize platform.
Feature Importance & Explainability with Arize
Are you correctly leveraging explainability in your ML models? Do you know the level of explainability that is needed for your specific use case?
June 28th
, 9:00am
-9:30am
PST
Read more →
Feature Importance & Explainability with Arize
June 28th
, 9:00am
-9:30am
PST
Learn how to leverage explainability, correctly, in your ML workflow. This hands-on workshop educates participants in the nature of explainability by demonstrating how teams across industries are using it to better understand their ML models. Afterwards participants will be able to apply explainability in their own models in production using the Arize platform.
Unstructured Data & Embeddings Tracking
If you are an enterprise looking to learn more about ML Observability for your team we have something you might be interested in.
July 5th
, 9:00am
-9:30am
PST
Read more →
Unstructured Data & Embeddings Tracking
July 5th
, 9:00am
-9:30am
PST
Are you able to track embedding drift to surface problems in your ML models? Do you know how to log embedding vectors and unstructured data with your model inferences?
Join us July 5th to learn how to troubleshoot your unstructured language models. As a company focused on building software to help humans understand how AI works, this hands-on workshop educates participants on tracking production models with embeddings — the core of how deep learning models represent structures, mappings and manifolds that are learned by models. Afterwards participants will be able to ingest and track embedding vectors representing their text data.
Fairness & Bias Tracing with Arize
Are you sure the ML models your company has in production are fair? Do you know if you are identifying model bias correctly and instantly in your current workflow?
July 12th
, 9:00am
-9:30am
PST
Read more →
Fairness & Bias Tracing with Arize
July 12th
, 9:00am
-9:30am
PST
Learn how to identify Fairness & Bias immediately in your current production workflow. This hands-on workshop educates participants in the nature of Fairness & Bias by demonstrating how teams across industries are using it to better understand their ML models. Afterwards participants will be able to identify Fairness & Bias in their own models using the Arize platform.
