DataRobot Enterprise AI Suite for AWS

My main use case for DataRobot  is that it is a platform at an enterprise AI level that every organization uses to build, deploy, and govern each machine learning model at scale. It is basically an experiment to build, monitor, and govern AI models and also recognize some leadership in AI governance and ML operations, with nearly half of the Fortune 500 companies using it.

I generally use DataRobot  in healthcare projects. We integrate DataRobot into our AWS  ecosystem, and it improves our patient healthcare through predictive analytics, resulting in fast diagnostics and better resource allocation. Efficiency in production and predictive maintenance are key aspects of my main use case with DataRobot.

The best features DataRobot offers include automated machine learning and AI collaboration tools.

The automated machine learning and AI features of DataRobot have helped us build predictive models rapidly using hundreds of algorithms. The AutoML has significantly supported our collaborative efforts, as we generally use this tool in a centralized workspace for data scientists, analytics, and business users. This collaboration enhances reliability, with extensive documentation and a realistic approach towards real data. Experimentability and audit documentation are also well managed in this feature.

DataRobot has positively impacted my organization by driving an AI platform that encompasses the entire AI lifecycle, helping us experiment, build, deploy, monitor, and govern AI models in a secure and scalable way. It is particularly realistic in healthcare as we focus more on faster diagnoses and patient resource allocation through predictive analytics.

DataRobot could improve by attaching more advanced AI features, which would empower its daily use to be more responsible, efficient, and provide real-time examples. This enhancement would demonstrate how AI can transform industries, cut costs, and drive innovations.

The integration of DataRobot would greatly benefit from allowing more realistic tools and would be improved if it integrates more comprehensively with AWS  cloud and other cloud platforms.

I have been using DataRobot since the last four years.

DataRobot is stable.

DataRobot's scalability is impactful, as it really helps maintain various solutions across different requirements and features. It is mainly used in healthcare, which contributes to its strong scalability.

The customer support for DataRobot is good, offering various options whenever support is needed, including licensing features.

I started with DataRobot itself and have not used a different solution.

My experience with pricing, setup cost, and licensing for DataRobot has been positive, especially regarding licensing. I am not sure about the costs since a different team handles that; my focus is more on the technical aspects.

I am not sure what my other technical team evaluated, but DataRobot was their first priority.

I have seen a return on investment as it has significantly helped us save time and improve reliability for different customers. It has enabled time-saving and cost-saving, which are key benefits of DataRobot.

I am satisfied with DataRobot.

Specific outcomes from using DataRobot include faster diagnosis times and improved resource optimization. It also assisted other teams, including those in finance with fraud detection and credit risk modeling.

DataRobot's security ensures compliance and regulations like GDPR and HIPAA, supporting all types of these regulations. This compliance and security aspect is crucial and it is scalable and flexible towards supporting any hybrid and multi-cloud environment.

Regarding accuracy and reliability of output, DataRobot has helped us process live time streams towards customer purchases and IoT sensors. It has effectively demonstrated how accuracy is maintained and has provided the right data to end users, thereby enhancing both accuracy and scalability.

I would suggest that anyone interested in an automated AI tool or ML tool consider DataRobot, as it provides a comprehensive AI platform suitable for enterprises at a scale level. It is ideal for those looking to utilize successful real-world stories from top industry levels to demonstrate how AI can transform industries, cut costs, and drive innovation. I have provided this review a rating of eight out of ten.

Hybrid Cloud

Amazon Web Services (AWS)

What do you like best about the product?

The interface is pretty easy to navigate. It also makes machine learning very accessible cutting down a lot of manual work.

What do you dislike about the product?

There honestly doesn’t seem much to dislike but sometimes the number of features can feel a bit overwhelming. The learning curve takes some time.

What problems is the product solving and how is that benefiting you?

It is definitely helping the with the manual work. Instead of spending weeks in writing codes, we are able to get accurate models in a short span of time.

What do you like best about the product?

I really like how quickly I can get useful information out of DataRobot. I can upload my data and start seeing results almost right away, which saves me a ton of time.

The automated modeling is a big help because it tries different approaches for me instead of requiring everything to be done manually. I also appreciate that I can still understand what’s driving the predictions, so it doesn’t feel like a black box.

What do you dislike about the product?

One downside is that it can feel a bit overwhelming at first. There’s a lot happening in the UI, and it isn’t always clear where to click or what to focus on, especially for new users. A more streamlined layout or a guided experience would make it easier to get started.

Performance can be another challenge. Some processes take a while, particularly with larger datasets. which can slow everything down and interrupt the workflow.

Overall, simplifying the interface, adding more transparency in certain areas, and improving speed would make a big difference.

What problems is the product solving and how is that benefiting you?

DataRobot addresses how slow and complicated it can be to build and use machine learning models.

For me, the biggest benefit is speed. I can go from raw data to something usable pretty quickly, rather than spending a lot of time testing things manually. It also makes it easier to get results without needing deep expertise, which helps me move forward with more confidence.

My main use case for DataRobot  is to perform predictive analysis and automation of machine learning workflows. I use it to quickly build, test, and deploy models without extensive coding. One of the examples is I use DataRobot  to predict which students are likely to accept the university offer. It basically helps us and the admission team to focus their efforts more efficiently. It also helps us with data matching and cleaning in large data sets, which reduces manual work.

The prediction helps our team and the admission team to prioritize outreach to the students who are most likely to accept the offer. They inform marketing and follow-up strategies as well, making efforts more efficient and quicker. One example is if DataRobot predicts a student has a high likelihood of accepting, the team can send personalized emails or call them to provide guidance and support directly to these students. It basically focuses on these specific students which have been just highlighted by DataRobot. It also reduces time spent on students who are unlikely to enroll, allowing us to use our resources more efficiently only on the people who we think are actually going to come back and enroll with us.

I do use DataRobot for many other things as well. For example, other than the target of student enrollment, I use DataRobot for data cleaning. I do the cleaning of deduplication as well. I also use this to detect any anomalies. It basically helps me to automate all the repetitive tasks and saves me some time. One example I can share is I use it to flag duplicate student records across multiple systems, which used to take us hours to do before, and now it's done a lot more quickly by using DataRobot.

There are many features that I appreciate about DataRobot. Some of the features which I personally prefer are the ones that save time. First, I would start with the automating features. If I want to do the data preparation, clean the raw data, or upload a student admission data, DataRobot automatically generates features such as the number of applications in the last month and previous offers accepted, and it can remove duplicates as well. This is one of my favorite features. Secondly, it tests my machine learning models for me, and the testing and selection are very efficient. For example, if I want to run many algorithms, DataRobot will compare them and pick the best one, saving me time from manually checking which one is the best. Lastly, another feature that I appreciate is the integration and scalability with our cloud system. It helps us connect with the various data sources we work with in our university, such as SITS, Azure  SQL, and CSV exports, allowing DataRobot to handle joins and feature engineering effectively without requiring extensive coding from me.

DataRobot has positively impacted our organization in many ways. First, it has improved efficiency; tasks such as model testing, feature engineering, and predictions that used to take us days or weeks can now be accomplished in hours. This ultimately helps us make better decisions, particularly with admission data where we can rely heavily on the predictions made by DataRobot. It has also helped to reduce a lot of manual work and has allowed me to execute automation tasks more quickly. Furthermore, DataRobot provides scalable analytics, enabling us to run multiple predictive models across different departments without needing extra staff or extensive infrastructure. For instance, it allows the admission team to prioritize outreach to students likely to apply, ensuring we spend our resources effectively.

Aside from the many advantages of DataRobot, I believe there are areas that could be improved based on my experience. There is a lack of transparency in the models; sometimes it feels like a black box. For example, when I uploaded a large data set of about two gigabytes for processing, the time taken was slower than expected. Additionally, the handling of bigger data sets could be better, as it performs extremely well with smaller datasets but can lag with larger ones. The integration with some other tools used in our organization can also be challenging, and more flexibility for custom pre-processing and advanced model tuning would be beneficial.

In terms of support and documentation, I believe improvements are needed. For instance, the response time from DataRobot could be quicker, which would be appreciated when we need assistance. The documentation is generally sufficient, but it can be lengthy and could use more real-world examples and step-by-step tutorials for better clarity. Lastly, creating a client community where users can share experiences and solutions might enhance the overall value and learning curve.

I have been using DataRobot for about one year, which is about the past 12 months.

DataRobot's customer support is good but could improve with quicker response times and better documentation or community support. The scalability is robust, managing large data sets, although it sometimes slows down when processing bigger data, but being cloud-hosted enables automatic resource scaling, which supports collaboration across teams.

Neutral

Previously, we used different solutions, including manual model building through Python, Excel, and Azure  ML for some projects. We switched due to the burdensome manual workflows that were time-consuming and required extensive coding, making it difficult to test multiple models quickly. DataRobot allowed us to experiment faster, achieve better model accuracy, and facilitate simpler collaboration without needing high-level programming skills.

We have indeed seen a return on investment. On average, we're saving about 10 to 15 hours per project. The efficiency has greatly improved; tasks that used to take a day now take mere hours. While we haven't reduced staff, their workload has lightened, enabling them to accomplish more within the same timeframe. The standout metric remains the 10 to 15 hours saved per project.

While pricing falls more under my IT colleagues, from my perspective, the overall experience feels justified. The premium pricing is reasonable for the value provided, and I'd say it's worth the investment. The setup cost was minimal because it's cloud-hosted, eliminating the need for heavy on-premises infrastructure, allowing us to start using it immediately after purchase.

We evaluated several options, including Azure Machine Learning  and manual Python workflows. DataRobot offered the best combination of automation, model accuracy, and ease of use, which ultimately saved us significant time and resources, making it the clear choice.

For those looking into DataRobot, I recommend starting with a small project to grasp the workflow before scaling. Utilizing the automations offered and dedicating some time for training is key, along with collaborating and sharing models and dashboards within the team to maximize the platform's value.

DataRobot is a powerful, user-friendly platform that saves time and provides accuracy, although improvements are needed in handling larger data sets and flexibility. You can use my real name for the public review. I have provided this review with a rating of 7.

Public Cloud

Amazon Web Services  (AWS )