Advanced Model Operationalization ★★★ Expert Level
This course will teach how you to operationalize your models in a scalable and robust way. You first learned how to put your models behind APIs in our Model Operationalization course, now advance those skills to deploy your APIs into the Cloud! This will enable your models to generate predictions for the whole business at any time, at any scale.
Currently there are no scheduled dates for this course. To be notified about upcoming dates, please choose "Reserve a seat".
We're sorry, but all tickets sales have ended because the event is expired.
*If you are a group of 5 or more, we are happy to accommodate a date for the training that suits you best. If so, please choose the "Reserve a seat" option.
Advanced Model Operationalization
About the courseOne of the most critical but often overlooked aspects of the machine learning process is the operationalization of models. Machine learning models have no tangible business benefits until they’re operationalized, and few businesses are prepared to integrate them into the real-life context of operations. This Advanced Model Operationalization course practices the specific steps to removing your AI initiative from the silos and putting it into production at scale. The first day will begin by teaching you how to deploy APIs that use models into the Cloud. Then we will look into serverless deployment, when and why to use it along with the pros and cons. To finish off we will delve into governance and control mechanisms for model operationalization including orchestration, versioning of models and reproducibility of results. The second day will have a strong focus on operationalizing models in Cloud machine learning services like SageMaker, understanding how to leverage this to train, deploy and improve models. The final stretch will push you to combine all this knowledge and design a solution for model operationalization. You will have to compare the different services and deployments methods and consider the business needs and technical requirements to tailor a solution that fits.
Why this is for youBeing able to build a model is not enough anymore, you also need to be able to bring it into production in a robust way. And moving from the academics of machine learning to the deployment of it is difficult. There’s a wall between the data science and the economics in projects. Months go by with experimenting, building, tweaking, and soon your stakeholders are getting impatient, thinking the final product will never appear. Thus, practicing the tools to operationalize your models is the most crucial skill for achieving impact. The tool this training focuses on, Cloud services, is gaining popularity rapidly to become a leading player with advantages in availability and scalability. Many companies are making the transition to the Cloud and you should be prepared with the knowledge and best practices too.
For whomThe content of this training hits right at the intersection of data science and data engineering. Therefore, this training is perfect for both Data Scientists and Data Engineers who have already completed the pre-requisite badge Model Operationalization (4210). If you currently feel overwhelmed by all the options and services available for operationalization and are seeking some clarity, this course will be your guide; helping you sort the advantages and disadvantages of each option, so you can make informed decisions and optimize model performance over time.
What you’ll learn
- Deploying APIs – Deploy APIs that use models for inference in the Cloud
- Serverless deployment – Explain what serverless deployment is and argue its pros and cons
- Governance and control – Explain and understand the importance of governance and control mechanisms for model operationalization
- Leveraging Cloud ML – Operationalize models in Cloud machine learning services like SageMaker and understand how to leverage the Cloud for model operationalization
- Architecting for model operationalization – Design a solution for model operationalization considering business needs and technical requirements