In a recent blog post, Google announced the beta of Cloud AI Platform Pipelines, which provides users with a way to deploy robust, repeatable machine learning pipelines along with monitoring, auditing, version tracking, and reproducibility. 

With Cloud AI Pipelines, Google can help organizations adopt the practice of Machine Learning Operations, also known as MLOps – a term for applying DevOps practices to help users automate, manage, and audit ML workflows. Typically, these practices involve data preparation and analysis, training, evaluation, deployment, and more. 

Google product manager Anusha Ramesh and staff developer advocate Amy Unruh wrote in the blog post: 

When you’re just prototyping a machine learning (ML) model in a notebook, it can seem fairly straightforward. But when you need to start paying attention to the other pieces required to make an ML workflow sustainable and scalable, things become more complex.

Moreover, when complexity grows, building a repeatable and auditable process becomes more laborious.

Cloud AI Platform Pipelines – which runs on a Google Kubernetes Engine (GKE) Cluster and is accessible via the Cloud AI Platform dashboard – has two major parts: 

• The infrastructure for deploying and running structured AI workflows integrated with GCP services such as BigQuery, Dataflow, AI Platform Training and Serving, Cloud Functions, and
• The pipeline tools for building, debugging and sharing pipelines and components.

With the Cloud AI Platform Pipelines users can specify a pipeline using either the Kubeflow Pipelines (KFP) software development kit (SDK) or by customizing the TensorFlow Extended (TFX) Pipeline template with the TFX SDK. The latter currently consists of libraries, components, and some binaries and it is up to the developer to pick the right level of abstraction for the task at hand. Furthermore, TFX SDK includes a library ML Metadata (MLMD) for recording and retrieving metadata associated with the workflows; this library can also run independently. 

Google recommends using KPF SDK for fully custom pipelines or pipelines that use prebuilt KFP components, and TFX SDK and its templates for E2E ML Pipelines based on TensorFlow. Note that over time, Google stated in the blog post that these two SDK experiences would merge. The SDK, in the end, will compile the pipeline and submit it to the Pipelines REST API; the AI Pipelines REST API server stores and schedules the pipeline for execution.

An open-source container-native workflow engine for orchestrating parallel jobs on Kubernetes called Argo runs the pipelines, which includes additional microservices to record metadata, handle components IO, and schedule pipeline runs. The Argo workflow engine executes each pipeline on individual isolated pods in a GKE cluster – allowing each pipeline component to leverage Google Cloud services such as Dataflow, AI Platform Training and Prediction, BigQuery, and others. Furthermore, pipelines can contain steps that perform sizeable GPU and TPU computation in the cluster, directly leveraging features like autoscaling and node auto-provisioning.

AI Platform Pipeline runs include automatic metadata tracking using the MLMD – and logs the artifacts used in each pipeline step, pipeline parameters, and the linkage across the input/output artifacts, as well as the pipeline steps that created and consumed them.

With Cloud AI Platform Pipelines, according to the blog post customers will get:

• Push-button installation via the Google Cloud Console
• Enterprise features for running ML workloads, including pipeline versioning, automatic metadata tracking of artifacts and executions, Cloud Logging, visualization tools, and more 
• Seamless integration with Google Cloud managed services like BigQuery, Dataflow, AI Platform Training and Serving, Cloud Functions, and many others 
• Many prebuilt pipeline components (pipeline steps) for ML workflows, with easy construction of your own custom components

The support for Kubeflow will allow a straightforward migration to other cloud platforms, as a respondent on a Hacker News thread on Google AI Cloud Pipeline stated:

Cloud AI Platform Pipelines appear to use Kubeflow Pipelines on the backend, which is open-source and runs on Kubernetes. The Kubeflow team has invested a lot of time on making it simple to deploy across a variety of public clouds, such as AWS, and Azure. If Google were to kill it, you could easily run it on any other hosted Kubernetes service.

The release of AI Cloud Pipelines shows Google’s further expansion of Machine Learning as a Service (MLaaS) portfolio – consisting of several other ML centric services such as Cloud AutoML, Kubeflow and AI Platform Prediction. The expansion is necessary to allow Google to further capitalize on the growing demand for ML-based cloud services in a market which analysts expect to reach USD 8.48 billion by 2025, and to compete with other large public cloud vendors such as Amazon offering similar services like SageMaker and Microsoft with Azure Machine Learning.

Currently, Google plans to add more features for AI Cloud Pipelines. These features are:

• Easy cluster upgrades 
• More templates for authoring ML workflows
• More straightforward UI-based setup of off-cluster storage of backend data
• Workload identity, to support transparent access to GCP services, and 
• Multi-user isolation – allowing each person accessing the Pipelines cluster to control who can access their pipelines and other resources.

Lastly, more information on Google’s Cloud AI Pipeline is available in the getting started documentation.

The post Google Announces Cloud AI Platform Pipelines to Simplify Machine Learning Development appeared first on Artificial Intelligence.

Google today announced the beta launch of Cloud AI Platform Pipelines, a service designed to deploy robust, repeatable AI pipelines along with monitoring, auditing, version tracking, and reproducibility in the cloud. Google’s pitching it as a way to deliver an “easy to install” secure execution environment for machine learning workflows, which could reduce the amount of time enterprises spend bringing products to production.

“When you’re just prototyping a machine learning model in a notebook, it can seem fairly straightforward. But when you need to start paying attention to the other pieces required to make a [machine learning] workflow sustainable and scalable, things become more complex,” wrote Google product manager Anusha Ramesh and staff developer advocate Amy Unruh in a blog post. “A machine learning workflow can involve many steps with dependencies on each other, from data preparation and analysis, to training, to evaluation, to deployment, and more. It’s hard to compose and track these processes in an ad-hoc manner — for example, in a set of notebooks or scripts — and things like auditing and reproducibility become increasingly problematic.”

AI Platform Pipelines has two major parts: (1) the infrastructure for deploying and running structured AI workflows that are integrated with Google Cloud Platform services and (2) the pipeline tools for building, debugging, and sharing pipelines and components. The service runs on a Google Kubernetes cluster that’s automatically created as a part of the installation process, and it’s accessible via the Cloud AI Platform dashboard. With AI Platform Pipelines, developers specify a pipeline using the Kubeflow Pipelines software development kit (SDK), or by customizing the TensorFlow Extended (TFX) Pipeline template with the TFX SDK. This SDK compiles the pipeline and submits it to the Pipelines REST API server, which stores and schedules the pipeline for execution.

AI Pipelines uses the open source Argo workflow engine to run the pipeline and has additional microservices to record metadata, handle components IO, and schedule pipeline runs. Pipeline steps are executed as individual isolated pods in a cluster and each component can leverage Google Cloud services such as Dataflow, AI Platform Training and Prediction, BigQuery, and others. Meanwhile, the pipelines can contain steps that perform graphics card and tensor processing unit computation in the cluster, directly leveraging features like autoscaling and node auto-provisioning.

AI Platform Pipeline runs include automatic metadata tracking using ML Metadata, a library for recording and retrieving metadata associated with machine learning developer and data scientist workflows. Automatic metadata tracking logs the artifacts used in each pipeline step, pipeline parameters, and the linkage across the input/output artifacts, as well as the pipeline steps that created and consumed them.

In addition, AI Platform Pipelines supports pipeline versioning, which allows developers to upload multiple versions of the same pipeline and group them in the UI, as well as automatic artifact and lineage tracking. Native artifact tracking enables the tracking of things like models, data statistics, model evaluation metrics, and many more. And lineage tracking shows the history and versions of your models, data, and more.

Google says that in the near future, AI Platform Pipelines will gain multi-user isolation, which will let each person accessing the Pipelines cluster control who can access their pipelines and other resources. Other forthcoming features include workload identity to support transparent access to Google Cloud Services; a UI-based setup of off-cluster storage of backend data, including metadata, server data, job history, and metrics; simpler cluster upgrades; and more templates for authoring workflows.

The post Google launches Cloud AI Platform Pipelines in beta to simplify machine learning development appeared first on Artificial Intelligence.

Google today announced the beta launch of Cloud AI Platform Pipelines, a service designed to deploy robust, repeatable AI pipelines along with monitoring, auditing, version tracking, and reproducibility in the cloud. Google’s pitching it as a way to deliver an “easy to install” secure execution environment for machine learning workflows, which could reduce the amount of time enterprises spend bringing products to production.

“When you’re just prototyping a machine learning model in a notebook, it can seem fairly straightforward. But when you need to start paying attention to the other pieces required to make a [machine learning] workflow sustainable and scalable, things become more complex,” wrote Google product manager Anusha Ramesh and staff developer advocate Amy Unruh in a blog post. “A machine learning workflow can involve many steps with dependencies on each other, from data preparation and analysis, to training, to evaluation, to deployment, and more. It’s hard to compose and track these processes in an ad-hoc manner — for example, in a set of notebooks or scripts — and things like auditing and reproducibility become increasingly problematic.”

AI Platform Pipelines has two major parts: (1) the infrastructure for deploying and running structured AI workflows that are integrated with Google Cloud Platform services and (2) the pipeline tools for building, debugging, and sharing pipelines and components. The service runs on a Google Kubernetes cluster that’s automatically created as a part of the installation process, and it’s accessible via the Cloud AI Platform dashboard. With AI Platform Pipelines, developers specify a pipeline using the Kubeflow Pipelines software development kit (SDK), or by customizing the TensorFlow Extended (TFX) Pipeline template with the TFX SDK. This SDK compiles the pipeline and submits it to the Pipelines REST API server, which stores and schedules the pipeline for execution.

AI Pipelines uses the open source Argo workflow engine to run the pipeline and has additional microservices to record metadata, handle components IO, and schedule pipeline runs. Pipeline steps are executed as individual isolated pods in a cluster and each component can leverage Google Cloud services such as Dataflow, AI Platform Training and Prediction, BigQuery, and others. Meanwhile, the pipelines can contain steps that perform graphics card and tensor processing unit computation in the cluster, directly leveraging features like autoscaling and node auto-provisioning.

AI Platform Pipeline runs include automatic metadata tracking using ML Metadata, a library for recording and retrieving metadata associated with machine learning developer and data scientist workflows. Automatic metadata tracking logs the artifacts used in each pipeline step, pipeline parameters, and the linkage across the input/output artifacts, as well as the pipeline steps that created and consumed them.

In addition, AI Platform Pipelines supports pipeline versioning, which allows developers to upload multiple versions of the same pipeline and group them in the UI, as well as automatic artifact and lineage tracking. Native artifact tracking enables the tracking of things like models, data statistics, model evaluation metrics, and many more. And lineage tracking shows the history and versions of your models, data, and more.

Google says that in the near future, AI Platform Pipelines will gain multi-user isolation, which will let each person accessing the Pipelines cluster control who can access their pipelines and other resources. Other forthcoming features include workload identity to support transparent access to Google Cloud Services; a UI-based setup of off-cluster storage of backend data, including metadata, server data, job history, and metrics; simpler cluster upgrades; and more templates for authoring workflows.

The post Google launches Cloud AI Platform Pipelines in beta to simplify machine learning development appeared first on Artificial Intelligence.

Gartner analysts have exhaled a “Magic Quadrant” report on Cloud AI developer services, concluding that while AWS is fractionally ahead, rivals Microsoft and Google are close behind, and that IBM is the only other company deserving a place in the “Leaders” section of the chart.

Gartner’s team of five mystics reckon that this is a significant topic. “By 2023, 40 per cent of development teams will be using automated machine learning services to build models that add AI capabilities to their applications, up from less than 2 per cent in 2019,” they predicted. The analysts also said that 50 per cent of “data scientist activities” will be automated by 2025, alleviating the current shortage of skilled humans.

The companies studied were Aible, AWS, Google, H20ai, IBM, Microsoft, Previson.io, Salesforce, SAP and Tencent. Alibaba and Baidu were excluded because of a requirement that products span “at least two major regions”.

AWS was praised for its wide range of services, including SageMaker AutoPilot, announced late last year, which automatically generates machine-learning models. However, some shortcomings in SageMaker were addressed during the course of the research, said the analysts. It is a complex portfolio, though, and can be confusing. In addition: “When users move from development to production environments, the cost of execution may be higher than they anticipated.” Gartner suggested developers attempt to model production costs early on, and even that they plan to move compute-intensive workloads on-premises as this may be more cost-effective.

Google was ranked just ahead of Microsoft on “completeness of vision” but fractionally behind on “ability to execute”. Gartner’s analysts were impressed with its strong language services, as well as its “what-if” tool, which lets you inspect ML models to assist explainability, the art of determining why a AI system delivers the results it does. Another plus was that Google’s image recognition service can be deployed in a container on-premises. Snags? The report identified a lack of maturity in Google’s cloud platform: “The organization is still undergoing substantial change, the full impact of which will not be apparent for some time.”

Microsoft won plaudits for the deployment flexibility of its AI services, on Azure or on-premises, as well as its wide selection of supported languages and its high level of investment in AI. A weakness, said the analysts, was lack of NLG (Natural Language Generation) services, though these are on the roadmap. The report also noted: “Microsoft can be challenging to engage with, due to a confusing branding strategy that spans multiple business units and includes Azure cognitive services and Cortana services. This overlap often confuses customers and can frustrate them.” In addition, “it can be difficult to know which part of Microsoft to contact.”

IBM is placed a little behind the other three, but still identified as having a “robust set of AI ML services”. Further, “according to its users, developing conversational agents on IBM’s Watson Assistant platform is a relatively painless experience.” That said, like Microsoft, IBM can be difficult to work with, having “different products, from different divisions, being handled by various development teams and having various pricing schemes,” said the analysts.

All four contenders can maybe take some comfort from Gartner’s report, which places the three leaders close together and IBM, with its smaller cloud product overall, not that far behind. Other considerations, such as existing business relationships, or points of detail in the AI services you want to use, could shift any one of them into the top spot for a specific project.

One of the points the researchers highlighted is that it can be cheaper to run compute-intensive workloads on-premises. Using standard tools gives the most flexibility, and in this respect Google’s recent announcement of Kubeflow 1.0, which lets devs run ML workflows on Kubernetes (K8s), is of interest. A developer can use Kubeflow on any K8s cluster including OpenShift. Google said it will support running ML workloads on-premises using Anthos in an upcoming release.

The post Gartner vision quest sees Microsoft, Google and IBM nipping at Amazon Web Services’ heels in cloud AI appeared first on Artificial Intelligence.