1. Microsoft Azure Machine Learning partner training
Server & Tools Business
4/28/2018
Microsoft Azure Machine Learning partner training
© 2012 Microsoft Corporation. All rights reserved. Microsoft, Windows, and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries.
The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

2. Modules MODULE 01: MODULE 02: MODULE 03: MODULE 04: MODULE 05:
Solution overview
MODULE 02:
Solution benefits
MODULE 03:
Selling Azure Machine Learning
MODULE 04:
Objections and compete
MODULE 05:
Licensing and pricing
MODULE 06:
Enabling partners with Azure Machine Learning

3. MODULE 02: Solution benefits
Welcome to module two of our Microsoft Azure Machine Learning (ML) partner training. In this module we will discuss the specific benefits the Azure ML solution brings to users and data scientists as well as the features of its core components that deliver those benefits.

4. Current state of the ML business
4/28/2018 7:59 PM
Current state of the ML business
No improvement in generations
Expensive
Expensive
Huge set-up costs—including tools, expertise, and computer/storage capacity—create unnecessary barriers to entry
The cloud changes the landscape
Siloed data
Siloed data
Siloed and cumbersome data management restricts access to data
Disconnected tools
Fragmented tools
Complex and fragmented tools limit
participation in exploring data and
building models
As we mentioned in our previous solution overview module, data science and machine learning have been around a long time. This begs the question as to why the technology is not being used more broadly. The answer boils down to two hurdles that have stymied many advanced technologies: expertise and money.
First, this work has been expensive. Our competitors are charging up to $100,000 per site license just to deploy the basics of a solution, which still does not speak to the time and cost of actual implementation.
Second, the data management side is highly cumbersome. Today, we see only a few companies that are actually doing data science work, but many of these are doing it only in individual departments, like finance, operations, or marketing. That means they have access only to siloed data and no connection to the data sources they need in order to get insight into the greater organization.
Third, even if they have access to organization-wide data, they are still often working in a vacuum when it comes to implementation. Industry-wide, ML modeling largely relies on one primary scripting language, called R, which is a robust platform but one that is usually completely unfamiliar to the rest of the development team because it is too specialized.
Fourth, even if organizations are able to solve all of these challenges, they then often reach a roadblock when it comes time to actually put their model into production. At this stage, models can either become stale because it took so long to develop them or they can grow old in production because the process to get them there was too time consuming to repeat.
But now there is a solution to all this and it is a key component of the Azure ML solution: the cloud. The cloud changes the game and the cloud is what Azure is all about.
Deployment complexity
Deployment complexity
Many models never achieve business
value because they encounter difficulty when deployed into production environments
© 2013 Microsoft Corporation. All rights reserved. Microsoft, Windows, and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

5. Microsoft Azure Machine Learning
4/28/2018 7:59 PM
Microsoft Azure Machine Learning
Built for a cloud-first, mobile-first world
Fully managed
No software to install, no hardware to manage; all you need is an Azure subscription.
Integrated
Drag, drop, and connect interface. Data sources with just a drop down run across any data.
Flexible
Built-in collection of best of breed algorithms with no coding required. Drop in custom R or use popular CRAN packages
Deploy in minutes
Operationalize models as web services with a single click; monetize in Azure Machine Learning Marketplace
Because Microsoft has developed our ML solution as a cloud service from the ground up, we are able to deliver a solution that addresses all the challenges we have just discussed and provide them in a way that is broadly accessible and budget friendly.
Most of you have heard us talk about our focus on enabling a cloud-first, mobile-first world. However, this should not be translated as abandoning the on-premises world, which is certainly not going anywhere, but instead putting the right tools in the cloud to enable modern business in ways that were difficult or even impossible before. ML is a great example.
Companies are generating and attempting to consume vast quantities of data today – more than ever before and this volume is still growing fast. Not only do companies have to think about ingesting that data and buying and managing all the systems necessary to gather and consume it, but they then have to think about buying and managing additional hardware and software for advanced analytics. For many companies, that kind of infrastructure burden is simply too heavy.
Azure ML solves this problem because it is a fully managed service in the cloud. That means customers need only an Azure subscription to utilize the full power of ML and advanced analytics. Just a subscription to gain access to a whole new landscape of data gathering, analysis, and insight, not roomfuls of infrastructure and big-ticket software licenses. Customers can even try it without a subscription because we offer a free version that only requires a Microsoft Live ID to get started right away. That means you can give your customers a hands-on demonstration of what ML and advanced analytics can do for their business without requiring any commitment on their part at all.
Another issue solved with the use of the cloud is the problem around integration. Azure data sources like HDInsight, Azure SQL Database, SQL Server in an Azure virtual machine (VM), and more can be brought into the modeling environment with a simple click on a drop down menu. And data that is not in the cloud can still be used in modeling by simply dropping what is called a “training set” (an initial set of source data) from on premises into the built-in storage space provided in Azure ML. The only data that must be in the cloud is the training set, but once the model is live as a web service it can be run across data in other locations.
We also understand, in terms of the talent base required for machine learning, that there is a wide range of skills we need to help with our service. That is why we have built a tool that is flexible enough to appeal to a user brand new to machine learning as well as the seasoned programmer who is experienced in ML modeling using industry standards, like R. By making available the market-tested algorithms we developed from our own businesses, including Bing, Xbox, Hotmail, and many more, a fledgling advanced analytics professional can drag and drop data sources in Azure ML Studio and experiment with machine learning without writing a line of code. More experienced implementers can drag and drop in their custom R code as well as use one of the over 350 R packages built into Azure ML. Essentially the use of Azure ML can be as simple or complex as your talent and business require.
Lastly, and this is a key point that customers need to understand about Azure ML, implementers can deploy a model as a web service in minutes. That is a unique capability in a market that is currently populated mostly be infrastructure-heavy and software-heavy on-premises solutions or solutions that purport to be cloud-enabled but only offer bits and pieces of everything needed to develop and fully deploy an ML model.
In practical terms, it means models that took weeks or months to put into production and apply to the business problem can now be applied right away. Additionally, once a customer has implemented an Azure ML web service they can begin gathering human input and applying it to their model without any additional development work. So for example, a senior marketing executive can call a database administrator or data scientist and ask them to adjust the output on their PowerBI dashboard to account for new variables as a result of a competitor’s movements. The implementer can now make that change and the dashboard can reflect it in hours – not days. That is something available only with Azure ML and it is an important differentiator for any customer.
© 2013 Microsoft Corporation. All rights reserved. Microsoft, Windows, and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

6. The Azure Machine Learning solution
Azure ML Studio
Browser-based
Designed for people without deep data science backgrounds
Supports deep science scenarios – R support, multiple models
Azure ML API
REST-based web service
Supports best-in-class algorithms
Reduces time from model experimentation to production
Azure Marketplace
Drag-and-deploy
Fast monetization of ML solutions and APIs
Quick source for free and third-party Azure ML APIs
We discussed this slide in our solution overview (module one), but for those that have not seen that session, these are the key components of the Azure ML solution. At its core, Azure ML is comprised of three things:
The Azure ML Studio
The Azure ML API
The Azure Marketplace
The Azure ML Studio is a big part of what makes Azure ML so valuable because it enables people without extensive experience in data science to take full advantage of ML analysis. It is an easily accessible browser-based solution with a simple and visual interface to let mainstream business workers access their ML data repositories for deep insight into how the business is functioning.
But that is not to say that Azure ML ignores advanced functionality or data scientists. All the cutting edge features data scientists need to build complex ML models are here, including support for R and deeper ML scenarios such as multiple model operations.
Another way for data scientists to get the most from Azure ML is by using the Azure ML API to create even more complex and dynamic models. The API is a REST-based web service that lets users extend the capabilities of Azure ML Studio with customization. But here, again, we have made it relevant to general users as well as data scientists by keeping it in the web service paradigm, which means it can also be accessible as an app plug-in or even using mobile apps.
The Azure Marketplace is the last core component of the Azure ML solution and a very important one. This is a one-click web service that lets users shop for ML components useful to their business or project and also allows data scientists to immediately market their expertise as web services.
But while Azure ML Studio, the Azure ML API, and Azure Marketplace comprise the core Azure ML solution, an important part of what makes Azure ML so competitive is that it integrates easily with the rest of the Microsoft Azure cloud data services portfolio. HDInsight, SQL Database, and Azure Storage and cloud infrastructure are all positioned to directly benefit the capabilities of Azure ML, which is a key part of what makes the solution flexible enough to address such a large variety of ML scenarios.
Azure cloud services
No software to install or infrastructure needed
Nearly unlimited file repositories via Azure Storage
Supports Azure data-related services – HDInsight, SQL Database

7. Azure ML Studio ML Studio New enhancements New experiment flow
Streamlined experiment page
New visualization for data tables
Browser-based environment supporting general users and data scientists
Immutable library of models including search, discover, and reuse
Wide range of features, machine learning algorithms, and modeling strategies
Ability to quickly deploy models as Azure web services to the ML API service
The heart of the Azure ML solution is Azure ML Studio. This is a browser-based integrated development environment (IDE) that uses drag-and-drop gestures and simple data flow graphs to let users that are not expert data scientists set up ML modeling experiments. For many tasks, users will not have to write a single line of code. ML Studio also features a library of time-saving sample experiments and sophisticated algorithms that have been developed by Microsoft Research, including the same algorithms Microsoft is using right now to run our Bing and Xbox cloud services among many others.
But Azure ML Studio also pays attention to the needs of more advanced users, like data scientists. It features a number of modules with which to build comprehensive predictive models, including state of the art ML algorithms such as scalable boosted decision trees, Bayesian Recommendation systems, Deep Neural Networks, and Decision Jungles developed at Microsoft Research.
We have made recent enhancements to Azure ML Studio, including a new and more intuitive experiment flow along with a new streamlined experiment page as well as new capabilities for visualizing data tables as customers make them part of an ML experiment.
Azure ML Studio is also critically important when it comes to deploying ML models and experiments. Using the Azure ML API, which we will talk about in a moment, Azure ML Studio can deploy ML web services as an integrated part of other analytical software and dashboards as well as making such services and algorithms available on the Azure Marketplace with just a few clicks. This is important because it lets data scientist customers monetize their experience immediately by allowing them to quickly expose their intellectual property to a broad swath of potential customers.

8. Azure ML API Azure ML Software Azure ML API Development Kit (SDK)
Custom data ingress and egress
Extends ML Studio with customization
Rich functionality – rules engine, R support, optimizer, simulation
Web-service and REST-based for easy creation and fast deployment
Allows general users and data scientists to run models as web services in minutes
Build apps that are easily accessible as web services, app plug-ins, or even mobile apps
Supports advanced data science, including R coding and 350 R packages included
What gives the Azure ML solution so much flexibility is largely the Azure ML API. This API allows customers to build powerful ML solutions, customize Azure ML Studio to their particular needs, and integrate Azure ML into other data analysis solutions and software.
And just like Azure ML Studio, the Azure ML API is accessible by users who are not sophisticated when it comes to advanced data analytics, but it also supports the needs of those who are. By enabling the API as a REST-based web service, we have made using it to run and publish models very easy. But by including richer functionality, including a rules engine, R support with 350 included packages, an optimizer, and simulation tools, we have also given it depth enough to address even the most advanced scenarios.

9. The Azure ML Marketplace
Published APIs available via search engine
One-click web service publishing allows data scientists to immediately monetize their expertise and creativity.
Immediate monetization in over 100 currencies
We have already mentioned the Azure Marketplace as the piece of the Azure ML solution that lets customers publish their ML services for fast deployment and lets data scientists monetize their efforts quickly and easily. The Azure Marketplace is an online store, but one devoted to products and services built with and for Microsoft Azure, including data services, virtual machines, apps, and, of course, machine learning services.
Using the Azure Marketplace, partners or their customers can access and leverage ML algorithms, plug-ins, or custom APIs. The right solution is quickly found using a search engine and can take the form of the software components just mentioned or full-fledged ML services for processes including forecasting, product recommendations, churn analysis, or social media analytics to mention only a few. Data scientists that develop these services for public consumption can publish them on the Azure Marketplace easily using Azure ML Studio, and immediately begin selling their solutions in over 100 currencies.
Selection of finished ML apps and APIs

10. 4/28/2018
© 2015 Microsoft Corporation. All rights reserved. Microsoft, Windows, and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries.
The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.
© 2012 Microsoft Corporation. All rights reserved. Microsoft, Windows, and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries.
The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.