Amazon SageMaker Valuable Features
SageMaker Studio sounds very interesting. Feature Store and data pipeline features are very interesting. The product is a one-stop shop. It allows people without much engineering knowledge to try out and deploy models in environments similar to the production environments. The tool makes our ML model development a bit more efficient because everything is in one environment. It is easy to manage compared to when things were in different components of AWS. Amazon SageMaker is in AWS, so I need not pay two bills. It is one less system to manage, so it is easier.
View full review »The product provides the ability to develop AI models relatively quickly. My team develops the models using the tool. We use AI quite extensively in our business. We use the tool for predictive analytics. It helps predict which trade might fail based on historical data. Automatic Model Tuning helps improve the productivity of the investment operation team. Typically, an analyst spends about 45% of their time collecting, organizing, and ingesting data. We were able to use the product to automate processes.
View full review »The Autopilot feature is really good because it's helpful for people who don't have much experience with coding or data pipelines. When we suggest SageMaker to clients, they don't have to go through all the steps manually. They can leverage Autopilot to choose variables, run experiments, and monitor costs. The results are also pretty accurate.
View full review »Buyer's Guide
Amazon SageMaker
May 2024
Learn what your peers think about Amazon SageMaker. Get advice and tips from experienced pros sharing their opinions. Updated: May 2024.
769,976 professionals have used our research since 2012.
We've had experience with unique ML projects using SageMaker. For example, we're developing a platform similar to ChatGPT that requires models. We utilize Amazon SageMaker to create endpoints for these models, making accessing them convenient as needed.
The main function I prefer in Amazon SageMaker is the ability to create endpoints for large models. I haven't explored features like Studio Lab yet, but I've found the tutorials very helpful. The platform is user-friendly, with documentation attached to everything, making it easy to navigate and learn. Overall, I especially like the Studio Lab feature.
In the Studio Lab, tutorials provide direct snippets for tasks like connecting to S3 from Amazon SageMaker. These standard snippets make implementation straightforward and simplify the development process for me.
View full review »The most valuable feature of Amazon SageMaker for me is the model deployment service. Serving the model is crucial because it seamlessly scales with the operation of the model, providing efficient infrastructure that adapts to the scaling needs, and ensuring optimal performance.
There is a lot of control in the solution over which terms you want to pick and choose. You don't have to pick the end-to-end machine learning operation solution. You can just choose deployment or training if you want, a benefit I saw in the solution.
You can leverage AWS's accelerated hardware to run your machine learning models, which is beneficial for improving a model's performance in terms of runtime, which is how long it takes to execute. Amazon SageMaker is an AWS product, and the company I work for already hosts all its services on AWS. Everything works well together if you're already on AWS.
The superb thing that SageMaker brings is that it wraps everything well. It's got the deployment, the whole framework. When you couple it with step functions, you can do some very powerful things. It manages deployment, you have model monitoring, and you have model quality checks. It's got a lot of end-to-end services one needs to get a full machine-learning pipeline running. While I say that I had a struggle and blame the product partially, I am also impressed with the ecosystem. I would still use it over and above other competing products, but I don't know the Google setup. I have worked very briefly with Azure, so I can't do a proper card-to-card comparison, but I do like the ecosystems AWS brings. If a client came along and asked me to set up a machine learning ecosystem, a full machine learning production deployment, I would use Sagemaker.
View full review »HJ
Harry Jais
Team lead at Assell
Feature Store, CodeCommit, versioning, model control, and CI/CD pipelines are the most valuable features in Amazon SageMaker.
View full review »The most valuable feature of the solution is Amazon SageMaker Canvas. The training and algorithm-based XGBoost modeling make it a good product for a startup, especially for companies that want to explore something but don't have a proper model. The instrument will be helpful for those who want to explore something.
The product aggregates everything we need to build and deploy machine learning models in one place. We log in to the cloud and have everything we need to build and deploy models.
View full review »SP
Sandeep_Pandey
Data Scientist at a tech vendor with 10,001+ employees
There are pre-built solutions for everything. For example, if you want to build a deep learning model, we already have AlexNet, the internet, and all of the packages are inside. You don't have to recreate the same thing from scratch, but instead, you can use their models. You can use their model and use their data, then you can use your data.
I am a big fan of their computational storage capabilities. It's a relational database itself. It's a new SQL and you get different types of services. That is one of the best things that I like when doing my research.
I cannot quantify it as it is based on your requirements, but I can say that it's very flexible and you are able to increase all of the RAM and the GPU support.
They are doing a very good job on their end. They are evolving. I have learned that they have already integrated an IDE into Amazon SageMaker. They are doing a good job of evolving.
View full review »JJ
Jaison Jose
Cloud Architect & Support Service Delivery Manager at Almoayyed Computers
The most valuable feature of Amazon SageMaker is that you don't have to do any programming in order to perform some of your use cases. As it is, we can start to use it directly.
View full review »SH
Srihari Hari
Solutions Architect at Emids
I am impressed with the tool's text extraction and its accuracy.
View full review »The most valuable feature of the solution is that it allows you to create API endpoints and that saves a lot of time for data scientists.
VK
VinayKumar9
Consultantconsultant at a tech services company with 1,001-5,000 employees
The most valuable feature of Amazon SageMaker is its integration. For example, AWS Lambda. Additionally, we can write Python code.
View full review »CD
reviewer1178424
Vice President & CIO at a logistics company with 201-500 employees
The most valuable features of this solution are the Random Cut Forest and the IDE.
View full review »PU
PankajUrmaliya
Lead Data Scientist at a tech services company with 201-500 employees
The deployment is easy and good. The documentation is pretty good also.
Integration with other AWS services is seamless.
View full review »Buyer's Guide
Amazon SageMaker
May 2024
Learn what your peers think about Amazon SageMaker. Get advice and tips from experienced pros sharing their opinions. Updated: May 2024.
769,976 professionals have used our research since 2012.