The Framework Way is the Best Way: the pitfalls of MLOps and how to avoid them
May 24, 2022 - Stefan Nica - 9 mins read
As our AI/ML projects evolve and mature, our processes and tooling also need to keep up with the growing demand for automation, quality and performance. But how can we possibly reconcile our need for flexibility with the overwhelming complexity of a continuously evolving ecosystem of tools and technologies? MLOps frameworks promise to deliver the ideal balance between flexibility, usability and maintainability, but not all MLOps frameworks are created equal. In this post, I take a critical look at what makes an MLOps framework worth using and what you should expect from one.
What's New in v0.8.0: Extend ZenML Any Way You Like!
May 18, 2022 - Alex Strick van Linschoten - 5 mins read
Release notes for the new version of ZenML. We've added a ton of extensibility improvements, documentation and guides that take away most of the hard work of figuring out how to add custom components. Our CLI also has been beautified and it should even run a bit faster too!
All Continuous, All The Time: Pipeline Deployment Patterns with ZenML
May 11, 2022 - Safoine El Khabich - 12 mins read
Connecting model training pipelines to deploying models in production is seen as a difficult milestone on the way to achieving MLOps maturity for an organization. ZenML rises to the challenge and introduces a novel approach to continuous model deployment that renders a smooth transition from experimentation to production.
What's New in v0.7.2 and v0.7.3: HuggingFace, Weights & Biases, LightGBM, XGBoost, and more!
April 28, 2022 - Hamza Tahir - 5 mins read
The release adds experiment tracking stack components, integrations with HuggingFace, Weights & Biases and also improvements to how the Seldon Core deployer handles secrets from the secrets manager component.
Predicting how a customer will feel about a product before they even ordered it
April 20, 2022 - Ayush Singh - 8 mins read
We built an end to end continuous deployment pipeline using ZenML for a customer satisfaction model that uses historical data of the customer predict the review score for the next order or purchase.
Podcast: Trustworthy ML with Kush Varshney
April 14, 2022 - Alex Strick van Linschoten - 1 min read
This week I spoke with Kush Varshney, author of 'Trustworthy Machine Learning', a fantastic guide and overview of all of the different ways machine learning can go wrong and an optimistic take on how to think about addressing those issues.
What's New in v0.7.1: Fetch data from your feature store and deploy models on Kubernetes
April 11, 2022 - Hamza Tahir - 3 mins read
The release introduces the Seldon Core ZenML integration, featuring the Seldon Core Model Deployer and a Seldon Core standard model deployer step. It also includes two new integrations with Feast as ZenML's first feature store addition and NeuralProphet adding to the growing list of training libraries supported.
'It's the data, silly!' How data-centric AI is driving MLOps
April 07, 2022 - Hamza Tahir - 9 mins read
ML practitioners today are embracing data-centric machine learning, because of its substantive effect on MLOps practices. In this article, we take a brief excursion into how data-centric machine learning is fuelling MLOps best practices, and why you should care about this change.
What's New in v0.7.0: 🔡 User Profiles and Secret Storage 🤫
March 28, 2022 - James W. Browne - 4 mins read
Release notes for the new version of ZenML. We've added the ability to store your stacks in profiles accessible system-wide instead of just in individual project folders. We also added a way to store and retrieve passwords and secrets, including the possibility to integrate the AWS Secrets Manager. Also: run individual steps on Vertex AI.
Run your steps on the cloud with Sagemaker, Vertex AI, and AzureML
March 25, 2022 - Hamza Tahir - 6 mins read
With ZenML 0.6.3, you can now run your ZenML steps on Sagemaker, Vertex AI, and AzureML! It’s normal to have certain steps that require specific infrastructure (e.g. a GPU-enabled environment) on which to run model training, and Step Operators give you the power to switch out infrastructure for individual steps to support this.
Podcast: Practical Production ML with Emmanuel Ameisen
March 18, 2022 - Alex Strick van Linschoten - 1 min read
This week I spoke with Emmanuel Ameisen, a data scientist and ML engineer currently based at Stripe. Emmanuel also wrote an excellent O'Reilly book called 'Building Machine Learning Powered Applications', a book I find myself often returning to for inspiration and that I was pleased to get the chance to reread in preparation for our discussion.
What's New in v0.6.3: Run Steps on Sagemaker and AzureML ☁️
March 14, 2022 - Alex Strick van Linschoten - 3 mins read
Release notes for the new version of ZenML. We've added the ability to run steps on AWS Sagemaker and AzureML. We added a new Tensorboard visualization that runs when using the Kubeflow orchestrator. You'll also find a lot of smaller improvements, documentation additions and bug fixes in this release.
How we made our integration tests delightful by optimizing the way our GitHub Actions run our test suite
March 9, 2022 - Alexej Penner - 16 mins read
As we outgrew our initial template Github Action workflow, here's the five things we added to our Github Action arsenal to fit our growing needs: Caching, Reusable Workflows, Composite Actions, Comment Triggers and Concurrency Management.
Podcast: From Academia to Industry with Johnny Greco
March 3, 2022 - Alex Strick van Linschoten - 1 min read
This week I spoke with Johnny Greco, a data scientist working at Radiology Partners. Johnny transitioned into his current work from a career as an academic — working in astronomy — where also worked in the open-source space to build a really interesting synthetic image data project.
How to painlessly deploy your ML models with ZenML
March 2, 2022 - Stefan Nica - 11 mins read
Connecting model training pipelines to deploying models in production is regarded as a difficult milestone on the way to achieving Machine Learning operations maturity for an organization. ZenML rises to the challenge and introduces a novel approach to continuous model deployment that renders a smooth transition from experimentation to production.
What's New in v0.6.2: ♻️ Continuous Deployment and a fresh CLI 👩💻
February 23, 2022 - Alex Strick van Linschoten - 4 mins read
Release notes for the new version of ZenML. We've added the ability to serve models using MLflow deployments, and we've refreshed how our CLI looks using the popular 'rich' library. You'll also find a lot of smaller improvements, documentation additions and bug fixes in this release.
Podcast: The Modern Data Stack with Tristan Zajonc
February 10, 2022 - Alex Strick van Linschoten - 1 min read
Tristan and Alex discuss where machine learning and AI are headed in terms of the tooling landscape. Tristan outlined a vision of a higher abstraction level, something he's working on making a reality as CEO at Continual.
How to improve your experimentation workflows with MLflow Tracking and ZenML
February 9, 2022 - Alex Strick van Linschoten - 7 mins read
Use MLflow Tracking to automatically ensure that you're capturing data, metadata and hyperparameters that contribute to how you are training your models. Use the UI interface to compare experiments, and let ZenML handle the boring setup details.
What's New in v0.6.1: Reach for the AWS and Azure Cloud! ☁️
February 7, 2022 - Alex Strick van Linschoten - 2 mins read
Release notes for the new version of ZenML. We now support AWS S3 and Azure Blob Storage as artifact stores. You'll also find a lot of smaller improvements and bug fixes in this release.
How to build a three-pointer prediction pipeline
February 2, 2022 - Alexej Penner - 10 mins read
We challenged ourselves to put our own tool to the test and set up a few pipelines to answer two questions: Did Steph Curry change the game of basketball? And how many three-pointers will be in the next NBA game?
Podcast: Neurosymbolic AI with Mohan Mahadevan
January 27, 2022 - Alex Strick van Linschoten - 1 min read
Mohan and Alex discuss neurosymbolic AI and the implications of a shift towards that as a core paradigm for production AI systems. In particular, we discuss the practical consequences of such a shift, both in terms of team composition as well as infrastructure requirements.
What's New in v0.6.0: whylogs integration and some big core architecture changes
January 26, 2022 - Alex Strick van Linschoten - 3 mins read
Release notes for the new version of ZenML. ZenML now supports whylogs (from whylabs) that logs data from your ML pipelines in production. We also made some sizeable (breaking 😢) core architecture changes
10 Reasons ZenML ❤️ Evidently AI's Monitoring Tool
January 21, 2022 - Alex Strick van Linschoten - 7 mins read
ZenML recently added an integration with Evidently, an open-source tool that allows you to monitor your data for drift (among other things). This post showcases the integration alongside some of the other parts of Evidently that we like.
Get to know ZenML through examples
January 6, 2022 - Alex Strick van Linschoten - 2 mins read
ZenML has a treasure-trove of examples available for users to get to know specific features. Using these examples, running them and pulling refreshed versions is easy with our CLI that takes on the heavy work for you.
