Azure Machine Learning Reviews & Product Details

Azure Machine Learning is an enterprise-grade service that facilitates the end-to-end machine learning lifecycle, enabling data scientists and developers to build, train, and deploy models efficiently. Key Features and Functionality: - Data Preparation: Quickly iterate data preparation on Apache Spark clusters within Azure Machine Learning, interoperable with Microsoft Fabric. - Feature Store: Increase agility in shipping your models by making features discoverable and reusable across workspaces. - AI Infrastructure: Take advantage of purpose-built AI infrastructure uniquely designed to combine the latest GPUs and InfiniBand networking. - Automated Machine Learning: Rapidly create accurate machine learning models for tasks including classification, regression, vision, and natural language processing. - Responsible AI: Build responsible AI solutions with interpretability capabilities. Assess model fairness through disparity metrics and mitigate unfairness. - Model Catalog: Discover, fine-tune, and deploy foundation models from Microsoft, OpenAI, Hugging Face, Meta, Cohere, and more using the model catalog. - Prompt Flow: Design, construct, evaluate, and deploy language model workflows with prompt flow. - Managed Endpoints: Operationalize model deployment and scoring, log metrics, and perform safe model rollouts. Primary Value and Solutions Provided: Azure Machine Learning accelerates time to value by streamlining prompt engineering and machine learning model workflows, facilitating faster model development with powerful AI infrastructure. It streamlines operations by enabling reproducible end-to-end pipelines and automating workflows with continuous integration and continuous delivery (CI/CD). The platform ensures confidence in development through unified data and AI governance with built-in security and compliance, allowing compute to run anywhere for hybrid machine learning. Additionally, it promotes responsible AI by providing visibility into models, evaluating language model workflows, and mitigating fairness, biases, and harm with built-in safety systems.

Seller

Microsoft

Discussions

Azure Machine Learning Community

Solution Type

All-in-One

Overview by

Don Scott (GM, AI Product @ Microsoft)

Value at a Glance

Averages based on real user reviews.

Time to Implement

3 months

Return on Investment

11 months

View More Pricing Information

Top-Rated Alternatives

Gemini Enterprise Agent Platform

View All Alternatives

User Insights

Average based on 90 real user reviews.

Implementation Time

3 months

Perceived Cost

$$$$$

Verified User in Accounting

Mid-Market (51-1000 emp.)

6/19/2023

"Its a best blend of all the tech we can use in machine learning"

5/5

What do you like best about Azure Machine Learning?

The environment is best we can use the environment veru efficiently Review collected by and hosted on G2.com.

What do you dislike about Azure Machine Learning?

Nothing The environment is best we can use the environment veru efficiently Review collected by and hosted on G2.com.

Verified User in Management Consulting

Small-Business (50 or fewer emp.)

8/12/2021

"A tool for non-programmers to contribute to data-analysis and (basic) machine learning pipelines"

2.5/5

What do you like best about Azure Machine Learning?

Having known the history of this too, which started with Microsoft's purchase of a local DC analytics R shop, "Revolution Analytics" a half decade ago, I think that the product has matured considerably since that time, not only in its functionality with respect to R, but now expanded to Python. The UI is crisp, and there are a lot of potential modules you can leverage in your analytics pipeline as you author the process from ingestion to output. Review collected by and hosted on G2.com.

What do you dislike about Azure Machine Learning?

Like SPSS Modeler (and Watson studio), these visual programmatic tools are still limited. You can't easily integrate "cutting edge" procedures which may not yet be even in modern repositories in this environment, and the drag-and-drop process of building your pipeline is so much slower than typing (unless you can't type decently fast). Further, it's easier to get started in the field by just programming since there are so many more help guides and questions/answers in online forums like StackOverflow than there are for a bespoke tool like this one. Review collected by and hosted on G2.com.

Recommendations to others considering Azure Machine Learning:

If you are thinking Azure as an augmentation to your data science pipeline concerning onboarding non-programming staff to your workflow, this may not work out well. There is a lot one can do with this tool, but it's still limited compared to all the new libraries and functionalities that being developed every day in third-party/open-source. Consider the cost-benefits of purchasing this service vs. hiring a data scientist with sound programming/engineering skills Review collected by and hosted on G2.com.

What problems is Azure Machine Learning solving and how is that benefiting you?

I used this tool for a limited amount of time as part of my access to Azure cloud. At first, we thought it could be an onboarding ramp for a data analyst to engage in machine learning pipelines Review collected by and hosted on G2.com.

Zayed R.

Programmer Analyst

Enterprise (> 1000 emp.)

8/19/2020

"Easy to build the machine learning model and deploy in the cloud"

4/5

What do you like best about Azure Machine Learning?

For a data scientist, Azure Machine Learning provides a numerous number of components available to create the model.

* Easy to deploy the model

* Easy to create the experiment Review collected by and hosted on G2.com.

What do you dislike about Azure Machine Learning?

Difficult to integrate the data for creating the model Review collected by and hosted on G2.com.

Uayeb C.

Advanced Analytics Manager

Small-Business (50 or fewer emp.)

10/31/2019

"Becoming in a data scientist"

5/5

What do you like best about Azure Machine Learning?

Azure machine learning is one of the best tools of Microsoft Azure to work predictive analytics, Data Scientist of world are researching methods to deploy their solutions. Azure Machine Learning Studios comes to solve this challenge. This tool has a command to commit new web services, those are easy to implement in different data pipelines technologies.

As well this great tool was designed for data scientists’ beginners, why? One of the biggest challenges for these new data scientists is code in programming languages. Coding, you must create your own functions to clean, transform and train models. Using azure ML you just need grad and drop modules and set parameters of what you expect as output.

Each one of those modules has as an output a Dataframe or predictive results, these results can easily be worked with python, R and SQL. Why this is an important feature? If you are an experimented data scientist, you will be able to use your own transformations functions. Review collected by and hosted on G2.com.

What do you dislike about Azure Machine Learning?

Azure machine learning studio right now does not have a way to scale the computer power, however, azure machine learning services you can set the power of the clusters. Another thing is that if you have just one work environment the main page just print all the experiment and it could be that you are working in different projects. Review collected by and hosted on G2.com.

Recommendations to others considering Azure Machine Learning:

Explore Azure databricks Review collected by and hosted on G2.com.

What problems is Azure Machine Learning solving and how is that benefiting you?

We use AMLS to connect with students, we process student information to predict who will dropout next academic term, we extract the information from the blob storage, what we are storing, academic, financial any kind of tabular information. After that we run all our custom transformations and merging with the native modules we are identifying the most important driver of student poor performance. Review collected by and hosted on G2.com.

José P.

Senior Data Engineer

Computer Software

Mid-Market (51-1000 emp.)

10/30/2019

"Machine Learning Processes Made Easy in the Cloud"

4.5/5

What do you like best about Azure Machine Learning?

- Azure Machine Learning Studio gives the user a numerous amount of components to tailor each individual process, covering most of the needs that one would have when developing these kinds of data procedures.

- Experiments that involve training a model can generate automatically a new experiment and web service that does the prediction job.

- Machine Learning Experiments can be deployed as a web service executable through an Endpoint.

- The free tier allows anyone to know how the tool works.

- It is easy to add new people to the environment and set the permissions necessary.

- Already provides code samples in different programming languages to execute created Web Services.

- Experiment creation is extremely intuitive as it has a drag and drop functionality. Review collected by and hosted on G2.com.

What do you dislike about Azure Machine Learning?

- Lacks a way of cleaning the final destination of the data before insertion. It can be worked around but the tool ends up being just for generating information.

- The interface where you see all of your experiments, projects, or web services rarely fits a screen resolution.

- Sometimes the execution of an experiment bugs and says it is still executing. Refreshing the page reveals it had finished a long time ago. Review collected by and hosted on G2.com.

Recommendations to others considering Azure Machine Learning:

Even though algorithms for machine learning are given by the tool, you need to understand what the situation at which they are best used is, otherwise you won't get the most out of Azure Machine Learning Studio. Review collected by and hosted on G2.com.

What problems is Azure Machine Learning solving and how is that benefiting you?

Machine Learning model training is easy with this tool. Cleaning data for training, sampling, predicting, and inserting information can all be done using Azure Machine Learning Studio. Review collected by and hosted on G2.com.

Verified User in Retail

Small-Business (50 or fewer emp.)

7/26/2020

"Easy handle machine learning"

3.5/5

What do you like best about Azure Machine Learning?

I like the most about Microsoft machine learning server it's very easy to start.And it is also flexible to add additional things on the fly so I really like the flexibility Microsoft machine learning server is providing. Review collected by and hosted on G2.com.

What do you dislike about Azure Machine Learning?

limitation to Microsoft technologies and less flexibility in terms of multi technology integration Review collected by and hosted on G2.com.

Arlindo P.

Mid-Market (51-1000 emp.)

12/23/2019

"Excellent for beginners in the world of machine learning."

5/5

What do you like best about Azure Machine Learning?

It is the best solution for beginners, as it uses a GUI for interacting and developing tasks that handle data and construct predictive models and calibrations. Review collected by and hosted on G2.com.

What do you dislike about Azure Machine Learning?

Calculation may take a long time to perform depending on the job you are trying to perform, so we sometimes wait until a predictive model results. Review collected by and hosted on G2.com.

Recommendations to others considering Azure Machine Learning:

Take advantage of Microsoft's extensive documentation and knowledge base to build the element you have to hold on to create your first solution without any problems. Review collected by and hosted on G2.com.

What problems is Azure Machine Learning solving and how is that benefiting you?

It recently asked me to make some kind of forecast based on data collected via a website survey to calculate the approval level of their customers and therefore try to predict the sales volume in the next few months. While searching for different options, I found Azure ML to be checked, and I started taking my first steps by comparing different regression and data transformations, after a while through the help guides and support pages, and I realized how good this tool is when I first started hanging from it. It has so many tools at its fingertips that it has a must. Review collected by and hosted on G2.com.

Elif S.

Researcher

Mid-Market (51-1000 emp.)

12/18/2019

"Perfect Machine Learning Tool For Engineers"

4.5/5

What do you like best about Azure Machine Learning?

I use this software to design my machine-learning algorithms. It enables me to create whatever I want with some predefined functions. It is very useful for a new learner. There are also some videos, comments, papers about this software. You can easily check and understand how to use. Review collected by and hosted on G2.com.

What do you dislike about Azure Machine Learning?

The way that the software gives error seems a little bit complicated. It is hard to catch some errors. It is not clear how to handle them. Some predefined functions and some machine learning tools are not clear. It also takes some more time to run my codes with respect to some relevant softwares. Review collected by and hosted on G2.com.

Recommendations to others considering Azure Machine Learning:

I recommend this software to anyone who is interested in designing a machine learning algorithm. It is very helpful and easy to use. I really recommend. Review collected by and hosted on G2.com.

What problems is Azure Machine Learning solving and how is that benefiting you?

I have been using this software to run my machine learning codes. It is very user-friendly software. It helps me a lot in terms of its easy-to-use property. Review collected by and hosted on G2.com.

Abhaya P.

Senior SDET

Mid-Market (51-1000 emp.)

12/17/2019

Business partner of the seller or seller's competitor, not included in G2 scores.

"Awesome predictive model"

4.5/5

What do you like best about Azure Machine Learning?

It provides a medium using which we can deploy our machine learning models and consume it as an end user. It is very easy to use and user friendly. I was very satisfied using this product and would recommend all my friends to use this product. I am able to write code in Pycharm using the libraries from Machine Learning Server. I can use the python libraries such as Numpy, Pandas in my model and deploy it Machine Learning Server. Review collected by and hosted on G2.com.

What do you dislike about Azure Machine Learning?

However, I have found some problems while using this product. Sometimes, the server hangs and you get stuck. You cannot use it for some time and it requires a restart. As a user, sometimes, it can be very frustrating. Some other times it can be very complex to use as you need to understand its terminologies. Review collected by and hosted on G2.com.

Recommendations to others considering Azure Machine Learning:

You won't face any issues in deploying your model in this server as it is very easy and customer friendly. You can create your own models using Machine Learning Server's inbuilt libraries and deploy it in the server and start consuming the model. It provides you accurate and precise results. And one more thing to keep in mind is to use its resources in a limited way else you can generate a big bill. So, use the resources judiciously. Review collected by and hosted on G2.com.

What problems is Azure Machine Learning solving and how is that benefiting you?

Using this, I am able to solve many problems and arrive at accurate results. I use the model to predict alerts which might get generated in the application at a particular time. I even use the model to study the trends of alert generation. When, I deployed my model on this server, I was able to get satisfactory results and show them on a graph which was very much helpful for my manager and customers. I will surely recommend this to my friends and colleagues. Review collected by and hosted on G2.com.