Amazon SageMaker Reviews

In terms of the tool's use case, if it is serverless, and if the compute involved is not too high, or if it is a PoC kind of a thing, and you want the microservices kind of architecture to be going and go for a pay as you go model, you can use the tool. With the tool, you know what is happening, so maybe you can cut costs by going with an on-premises model and having a stable system for computing.

The models that the tool has, the libraries, and the ML libraries are rich. That is good, and that is one of the features of the tool. The other feature is how the tool interacts with other components of AWS, like Lambda and S3. It's seamless if you are using AWS architecture.

I was working with Amazon SageMaker Ground Truth. The serverless feature is important and worth it because you don't have to spin up an EC2 instance every time or a server. For us, the main attraction is that it's serverless. You pay only for the compute that you use, and then there is its ecosystem. Whenever you use something in AWS, the ecosystem is very rich.

For any cloud provider, the cost has to be substantially reduced, especially in the case of Amazon SageMaker, which is extremely expensive for huge workloads. In EC2, you have spot instances that cut costs tremendously, but you don't have that in Amazon SageMaker. You pay for the local usage. I would like to see better integration with GPUs. GPUs are very expensive for AWS or any cloud provider. NVIDIA has introduced options with Databricks for GPUs, so it would be interesting to see how Amazon SageMaker can parallelize GPU usage. I haven't used it to scale multiple GPUs automatically for model training. The key points are the cost and how effectively they integrate GPUs into the workload for training machine learning models. We want to see how seamless it is and how it can work. I haven't used multiple GPUs scaled automatically. For model training, the first concern is cost, and the second is how effectively they want to integrate GPUs into the workload for training machine learning models.

I have six to seven years of experience with Amazon SageMaker.

After production, a different team handles the tool. I would be there till the PoC phase and then after production we move on to other projects. From my standpoint, I rate the solution's stability a nine out of ten. I don't continuously maintain production workloads, as it is managed by a different team.

The tool is extremely scalable. If you go on and use it for the entire life cycle of ML, it's very expensive. I say it's scalable, but it's expensive. For a production project, I would think that deployment and inferencing can be in SageMaker. We will have to, at some point, move other stuff to a less costly thing. It is very scalable.

My clients are mostly mid-level to enterprise businesses. I have worked with huge clients, but our clients may not be that huge. When we worked as a product engineering partner to AWS for two years, I dealt with huge clients, but that was very specific. I was working with Amazon SageMaker Ground Truth. I have worked with Nissan, Sony, and Samsung, but even though it was ML, it was a very niche kind of thing. We were doing labeling support for the ML models and training data. We were doing training data and labeling by partly using ML. Sometimes, we used to use Amazon SageMaker as well, but that was very niche. If we were to embark on a complete ML journey with some clients, then I would say that I have dealt with small to mid-scale customers.

The technical support of the tool was good. I rate the technical support a ten out of ten.

My company takes a tool-agnostic approach and utilizes a variety of AI and ML tools across the ecosystem, focusing on Amazon, Microsoft, Databricks, and Snowflake.

Speaking about the product's initial setup phase, I am in an Amazon ecosystem. I find the tool good because it depends on whether you are very command-line or user-interface-driven. If you are user interface-driven, I think Azure is good. I am okay with the command lines and shell kind of an interface. I am more used to AWS than other cloud providers. I think it's good for me. For a programmer, it's easy. As a technical person, it's easy. As a business person, I think Azure is a little easier than AWS.

The tool is deployed on a private cloud. Sometimes, we don't use it ourselves and then just give access to the apps to the client, meaning we host it. We host the model, and then the clients use the model but don't have access to infrastructure. We mostly used AWS along with some other cloud providers too.

The tool's deployment was fast, and it took a week.

For the deployment, whatever we do in general for on-premise, we do almost the same thing in SageMaker. Just that it offers parallelism. We do EDA and all that. We first have the data in the S3 bucket, and then we do the EDAs. Then we do the training. Then, we do the grid optimization. First, we mostly start with the pilot. When I say that deployment takes one week, it's just for having it up and running and showing some results with the model. It's not for the complete production model, as it is a different cycle. We start the PoC with the pilot model first, and then we kind of do it. Whatever steps are involved for on-premises, we do the same steps for SageMaker. Just that the computing cost is less, and we don't have to spin up a server. For the rest of the steps, like getting and cleaning the data, we don't do a lot in SageMaker. We kind of do it outside, or we do it with EC2 instances or use PySpark, maybe, and Databricks because that is a little easier compared to SageMaker libraries. You can do it, but it's kind of expensive. We do all that data transformation, and then we do the modeling. Most of the time, 100 percent deployment is done in SageMaker, but the rest of the parts are a mix of technologies. There is an influenced pipeline.

The cost offers a pay-as-you-go pricing model. It depends on the instance that you do.

I used Amazon SageMaker as a customer from the client side, though I have worked as a contractor with Amazon for two years. I work for Persistent, but the engagement with our client was over last year. When I was working at Amazon for two years, I was working at Amazon. Now, we are partners as well. We are strategic partners for Amazon, Microsoft, Google, Databricks, Snowflake, and most of the ecosystems.

If you want to try out something with, say, for instance, you want to do a PoC for testing more, I mean, say, for instance, that you're doing a data annotation project. You want to see how it goes. You don't want to invest a lot, and you want to try it out and see whether it works or not. For those kinds of typical PoC situations, I would say Amazon SageMaker is good. You are using a microservices architecture, and you want to go serverless. That is the first use case. The second use case is that you want to go serverless and plug and play a lot of components rather than having a bulk of computing like EC2 and all that. You would rather have an Amazon setup that is serverless.

We use it for tuning, but it's just like any other tool, except for the fact that it's serverless. It's not that it significantly boosts anything; it's just a choice. Either we tune it on-premise or we tune it on the cloud. We use Azure, AWS, and all that. So, in terms of tuning, it's not special. It's just the way you tune any model in any environment, and that is not a huge thing. It is a good tool that works well with its components and other components. There's nothing special about the tuning itself. You can either use PySpark or other cloud technologies. It's not that we get a huge boost just because it's AWS.

The serverless feature and the complete lifecycle that can be handled inside SageMaker are important. It covers everything from training the model to deploying it and sometimes using it for data pipelines. However, we generally don't use it for pipelining and data transformations because it's expensive inside SageMaker. We do use it for model training, although sometimes we train outside. We also utilize model training and Amazon SageMaker JumpStart, which is pretty handy because you don't have to train the model from scratch. You can use it, especially for LLM settings, right out of the box. There are models inside SageMaker that make it a little faster, both from a computing perspective and from a bandwidth deployment perspective, so you don't have to spend a lot of time training before deployment. Amazon SageMaker JumpStart is definitely valuable, along with the whole lifecycle for ML as well.

I would recommend others if they want to do a quick PoC workload, or proof of concept, and if they want to do something very quick, then I would definitely recommend it. If it's a very huge production workload, then I might want to consider other options. But for anything where there is a PoC kind of thing, I would recommend products in such areas.

Speaking about AI, I can say that it's kind of quick to set up and get it running. I can't say specifically. We have worked on a lot of projects. We have worked with document processing projects a lot. In those cases, if you were asking about specific projects, I can remember a recent project where we were trying to digitize documents using manual annotation and automated ML models, but there, we didn't use SageMaker. We used Amazon SageMaker Ground Truth, which is under the umbrella of SageMaker. If you use Ground Truth, it's a SageMaker product. We were using SageMaker Ground Truth, which is pretty handy because it sits well in the environment. If you are specifically asking how it accelerated the process, it was easy to set up, and we just got going in less than a week. So, yeah, I can think of that example.

I rate the tool an eight out of ten.

We use the product for deploying machine learning models. We use it for the machine learning model development process.

We're currently implementing a project on a cross-selling model. It is like a standard XGBoost model. I’m evaluating the tool to see whether it will improve the workflow.

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.

The product must provide better documentation. I don't see a lot of documentation, particularly on the Studio feature. In general, there is not a lot of information about how to use Feature Store. I can see it there, but the documents are not very explanatory.

I have been using the solution for two years.

The product is stable. I rate the stability an eight out of ten.

The product is very scalable. I rate the scalability an eight out of ten. Six people use the product in our organization. We are planning to increase the usage.

We are premium AWS customers.

Positive

I have used Databricks before. From a feature perspective, Databricks is better than SageMaker. The user experience of Databricks is much better. SageMaker’s advantage is its cost. The data is already on AWS S3. It’s less of a hassle. SageMaker is convenient. Databricks is like a MacBook. I can still work with SageMaker, but it is not as pretty, and the flow is not as natural as on Databricks.

The setup is not very easy. I rate the ease of setup a seven out of ten. The deployment takes ten minutes.

The pricing is comparable. It is not very cheap. I rate the pricing an eight out of ten. The main reason why we're using it is because of its cost. We are aiming at keeping the costs at $100 per month.

The product can scale model training and deployment. It is one platform. It is easy to use. People who want to use the product must first focus on defining the workflow of their team without any tools and then see how the product adapts rather than trying to use all the features of the system. It can confuse us. Overall, I rate the solution a seven out of ten.

I use SageMaker to use a "bring-your-own-model" setup. For SageMaker AutoML, we're fine and happy with it. It is restricted because you can't move through multiple algorithms. It seems to work only with two. One of the things I am doing is prototyping, and it's proving quite difficult to get our model working how we want it to. It's proving complex with many moving parts, and the documentation is only partially helpful. SageMaker requires a lot of work to get it working.

I spent the last four months trying to get a prototype working and exploring to bring in a model while exploring alternate models and making prototypes work. We've stepped back to AutoML for now. We might be using EKS, so we bring our containers. Within the containers, we can work with what we need to work with.

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.

The solution is complex to use.

Some additional functionality would be for them to provide sample end-to-end card formation templates and try to unify the setup. At the moment, as you move from one set of documentation to the next, some of the documentation is for bringing your model, some of the documentation is for SDK, some of it's for API, some of it's for command lines, and some of it's for step functions. None of the documentation seems to be end-to-end. There are gaps in the documentation. It proves to require a lot of digging from the user to figure it out. I did get through the AWS machine learning specialty certification, but that proved to be a bit superficial. Though it covered a lot of ground, it didn't have the detail one would need for Sagemaker. I am self-taught in a lot of the stuff. I could dive deeper into some code and take time to get examples running. But I was consulting a startup, and they needed to move quickly.

I was hoping SageMaker would be easier to work with because I was expecting there would be examples we could repurpose that were more complete.

The new functionality I'd like to see is Amazon tuning attention to the documentation sets and the templates.

I've been using Amazon SageMaker for about two and a half years.

Stability is not a relevant metric anymore because SageMaker runs on its own underlying AWS serverless infrastructure, which is 100% reliable.

Folks better than me with more extensive resources and time have run and checked SageMaker ten times a second every second for 1,000 hours to see if they got a drop, but they haven't. It's serverless and bulletproof.

I rate the scalability a ten out of ten. Part of the problem is that AWS has limited the functionality of SageMaker in many ways to make it scalable. So it's scalability first and then functionality second.

The solution works well for medium-sized businesses and up. But even for small businesses, you can do some simple and quick elastic endpoints and get going quickly. The problem is the amount of work it takes for people to know what they're doing with Sagemaker, and those people are probably rare. I've been able to get things up and running in most cases in all sorts of AWS services, but I'm struggling. Small, medium, and large enterprises could use SageMaker with an automatic model, but it depends on the people's skills doing the deployment. A small business probably couldn't afford contractors, consultants, or data scientists. It's not about AWS. It's a problem with classic data science skills.

I rate the initial setup a three or two out of ten because it's very complex.

You don't pay for Sagemaker. You only pay for the compute instances in your storage. SageMaker is free.

I've had a little bit of a look at Azure, but I didn't get into the level of detail I did with Sagemaker. I have worked reasonably intentionally with DataRobot and H20. But SageMaker is a way bigger, way more capable platform. The AutoML is very simple, and it is much, much cheaper. The cost of SageMaker is nothing. By contrast, if you're using DataRobot, you'll pay $100,000 plus for a five-year license.

Anyone doing on-prem at the moment for anything but their core datasets or legacy systems that can't be moved is just paying useless money.

I rate Amazon SageMaker a seven out of ten. I'd recommend it to other users. It's worth syncing the time and effort into getting it running.

The primary use case of Amazon SageMaker is for training a small AI module for learning purposes. It was used for the training of a small machine learning model.

I appreciate the ease of use in Amazon SageMaker. I have not explored many features, as I am not deeply involved yet, but I aim to enhance my skills in the future. 

Based on my existing experience, it was straightforward to train a small machine-learning model. I initially used AWS Skill Builder for guidance, making it manageable without encountering challenges. 

In scalability, I found it highly scalable, having used the Jupyter notebook and other tools. By scaling the model, I've had a positive experience.

I would recommend having more walkthrough videos and articles beyond AWS Skill Builder. There should be additional articles within the services.

In terms of stability, I have not experienced any breakdowns. Although I have heard reports that it might break, I have personally never faced any issues with the stability of Amazon SageMaker.

I found that Amazon SageMaker is highly scalable. I used the Jupyter notebook and explored other available tools, which were useful depending on what I utilized. I plan to enhance my model in the future, which will allow me to share more about scalability once I fully scale the models.

The support team of Amazon is excellent. I found them to be very good.

Positive

I have not used anything else in cloud computing. Amazon SageMaker was my entry-level experience in cloud computing.

I would give Amazon SageMaker a solid score of eight out of ten since I have not used much of its services. 

Based on the small model I trained, I would recommend it to others. I have already recommended it to some colleagues, batchmates, and fellows. My advice to newcomers would be to look for walkthrough videos and articles to aid in their learning.

Our primary use case is to build machine learning models and manage them using Amazon SageMaker. We use various tools provided by SageMaker, such as the studio for machine learning, pre-processing data using Wrangler, and deploying models. Some users prefer using Jupyter notebooks for their own libraries while others use features like Jumpstart or Autopilot.

One of the most valuable features of Amazon SageMaker for me is the one-touch deployment, which simplifies the process greatly. Additionally, I appreciate the flexibility that the notebook provides, as it allows me to experiment with scripts. 

The solution offers the SageMaker Model Monitor, which helps monitor deployed models for performance issues like bias and drift. Also, the high scalability of SageMaker allows us not to worry about the underlying infrastructure, as it automatically adjusts based on demands.

The dashboard could be improved by including more features and providing more information about deployed models, their drift, performance, scaling, and customization options.

I have been working with Amazon SageMaker for about three years.

The stability of the solution is generally about an eight out of ten. Most instabilities arise from initial configuration errors rather than the infrastructure itself. Ensuring that the correct setup is chosen from the start minimizes these issues.

Amazon SageMaker is highly scalable, rated ten out of ten. It can scale up according to the demands detected by CloudWatch, providing a seamless experience without needing to manage the underlying infrastructure.

The customer service and support are rated as a five out of ten. The level of support depends on whether we are a premium AWS customer or not, with premium customers receiving better and more immediate support.

Neutral

The initial setup of SageMaker can be challenging for beginners, rated as a six, but easier for those with a background in machine learning, rated as a nine out of ten. Experience with machine learning is crucial for a straightforward setup. Without it, understanding the roles of different features can be a stumbling block.

Pricing is rated as a six, which is slightly more expensive compared to the budget yet adequate for the capabilities provided. On average, customers pay about $300,000 USD per month.

On a scale from one to ten, where ten is the best, I rate Amazon SageMaker as a nine. For new users evaluating SageMaker, it is important to remember that it takes some learning, however, the solution is straightforward and beneficial. There are no special prerequisites except having an account.

We use the solution to extract financial information and contractual data from unstructured documents.

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.

The solution requires a lot of data to train the model.

I have been using the solution for the past 12 months.

The tool’s stability is pretty high. I rate the stability a nine and a half or ten out of ten.

The scalability is very high. I rate the scalability a nine out of ten. We are a small team of AI analysts. We have half a dozen users.

The support is usually pretty responsive. The solution has a fair bit of content online. We haven't had any support challenges.

Positive

We were using Azure’s tool before. We switched to Amazon SageMaker because it allows us to sell it to larger institutional clients. AWS is more prevalent in the broader institutional segment.

The initial setup is relatively straightforward. The same team developing the models deploys and tests the solution. The tool requires a bit of ongoing maintenance. It is relatively easy to do.

The product is expensive. I rate the pricing a five or six out of ten.

We are partners and resellers. Overall, I rate the product a nine out of ten.

I am currently using SageMaker for a range of tasks including data cleaning, data visualization, exploratory data analysis, data labeling, training, and deploying models. I am also using it for machine learning projects.

Amazon SageMaker has accelerated our machine learning development, improved scalability, optimized resource use, and reduced costs, enabling us to deliver faster, more efficient, and scalable ML solutions.

SageMaker is a comprehensive platform where I can perform all machine learning activities. Specifically, the notebook feature is beneficial because it allows me to do everything from data cleaning to model training, dataset transformation, and data visualization on a single platform.

I had to create custom templates for labeling multi-data sets, such as text and images, which was time-consuming. It would be appreciated if SageMaker made it more adaptable for different types of datasets instead of just specific ones.

There have been no performance or stability issues with SageMaker.

SageMaker is scalable.

I have not used customer support services. However, I did receive notifications regarding resource usage, such as a restart notice during long computations.

Positive

I have worked with generative AI, LLM models, APIs from Google, GPT, and Llama.

I switched because Amazon sagemaker comprises of most activities I needed from other different solutions. 

The initial setup was very straightforward with a left-side panel aiding in easy navigation. It was quite simple and user-friendly.

If you are not on the free tier, SageMaker might be expensive, especially if you forget to shut down running applications. However, since I am using some free-tier services, cost efficiency is not a concern for me.

I have worked with other solutions such as generative AI models and APIs from other providers. I have used Jupyter notebook via Anaconda, VSCode and Google Collab for my day to day machine learning tasks.

I would recommend SageMaker because it serves as a one-stop solution where everything needed for machine learning can be done in a single platform, without looking for other platforms to perform individual tasks.

I'd rate the solution nine out of ten.

We use Amazon SageMaker for a specific project involving an airline company to analyze historical flight data. Initially, we created a proof of concept locally and then developed a data pipeline to enhance the flow for large data models. This project benefited from SageMaker's ability to handle large data with TensorFlow, leading to a more efficient MLOps process.

SageMaker has streamlined the process for us, particularly in handling TensorFlow and making data management auto-scalable. This has allowed for a more efficient process, especially when discussing with DevOps about MLOps capabilities.

The most important feature is the ease of controlling costs with pricing calculators, which is crucial due to our limited budget. The low-code MLOps aspect simplifies data science work and makes it accessible for engineers without a DevOps background. It's user-friendly for business teams as they can understand many aspects through the AWS interface.

While not specific to SageMaker, I've observed a trend where companies are more attracted to the marketing of Gemini and GPT, leading some to migrate their pipelines to these platforms. Amazon might need to emphasize its capabilities in generative models more effectively.

SageMaker has proven to be stable with only one issue in our history, which is impressive.

Before SageMaker, pipelines were created manually. Since using SageMaker, I've been impressed with its high scalability, rating it a ten out of ten.

We have used free support tickets, and the response time is normal, usually in one or two days.

Positive

The initial setup was not complex, taking us about one month to fully understand and set it up, with the actual setup time being around one to two weeks.

We managed the implementation with two people focusing on understanding and making additional tests.

Compared to other top cloud services like Azure and Google Cloud in Brazil, AWS pricing is competitive and merits a rating of five out of ten.

We are migrating some models to Azure to evaluate its GPT enterprise version.

It's important for new users to clearly define the testing process and understand what they aim to achieve. SageMaker facilitates a fast, secure, and scalable MLOps process.

I'd rate the solution nine out of ten.

Amazon SageMaker is a collaborative tool for our data science projects. It allows us to integrate efficiently, write and review code, and access all the necessary project tools.

Amazon SageMaker is highly valuable for managing ML workloads. It connects to AWS cloud resources, making it easy to deploy algorithms and collaborate using tools like GitLab. It offers a wide range of Python libraries and other necessary tools for modeling and algorithms.

Amazon SageMaker can make it simpler to manage the data flow from start to finish, such as by integrating data, usingthe machine, and deploying models. This process could be more user-friendly compared to other tools. I would also like to improve integration with Bedrock and the LLM connection for AWS.

I have been using Amazon SageMaker for the past two years.

I've never encountered issues with SageMaker's scalability. AWS provides all the necessary resources in terms of power and capacity.

The initial setup is straightforward. We have a team from the infrastructure department that ensures the system runs smoothly. The data science team also plays a role in monitoring the effectiveness of the models. The deployment process usually takes two to three months for the whole project, with various strategies involved. SageMaker integrates well with AWS features, and when deploying, I typically set up APIs to make the model accessible to other systems and connect it with GitLab for easier model control.

In terms of pricing, I'd also rate it ten out of ten because it's been beneficial compared to other solutions.

I would rate Amazon SageMaker a nine out of ten because while it has all the necessary features, there could be improvements in making the data flow more manageable.

The primary use case for Amazon SageMaker is leveraging its compute power, particularly for tasks like securing LMM notebooks using node instances. Additionally, its GPU capabilities are valuable for executing large language models. Users can create endpoints and access them from anywhere as needed.

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.

In my opinion, one improvement for Amazon SageMaker would be to offer serverless GPUs. Currently, we incur costs on an hourly basis. It would be beneficial if the tool could provide pay-as-you-go pricing based on endpoints. 

In the three months I've been using it, I've noticed that higher GPU instances can be quite costly. To mitigate this cost impact, serverless GPUs would be beneficial.

I have been working with the product for three months. 

I rate the solution's stability a nine out of ten. 

I rate the tool's scalability an eight out of ten. No issues with scalability as long as we ensure we have the necessary quotas in place before implementing a scalable process. I needed to request quota increases for certain services beforehand, and once those were provided, I could adjust the main and max nodes accordingly based on our planned requirements. My company has 25 users. 

We can schedule a direct call with the support team. 

Amazon SageMaker's Studio Lab feature differentiates it from products like Azure ML Studio. With Studio Lab, I can directly interact with the environment, making navigating and accessing documentation easier. In contrast, finding documentation and navigating Azure ML Studio was challenging.

However, we also use Azure for the Azure OpenEdge service, which operates on a pay-per-minute token basis. This payment model is not available in Amazon SageMaker. 

The initial setup and deployment process for Amazon SageMaker is straightforward. The only complexity I encountered was gaining access to the needed resources, which relied on coordination with the DevOps team. Once I had access sorted out, implementing my ideas for large language models and other models was comfortable. 

The tool's pricing is reasonable. 

I rate the overall solution an eight out of ten.