IBM Cloud Pak for Data Reviews

We are using IBM Cloud Pak for Data for the development of models, such as credit scoring, anti-money laundering, and fraud control cases.

The most valuable feature of IBM Cloud Pak for Data is the Modeler flows. The ability to develop models using a graphical approach and the capability to connect to various sources, as well as the data virtualization capabilities, allow me to easily access and utilize data that is dispersed across different sources.

The interface could improve because sometimes it becomes slow. Sometimes there is a delay between clicks when using the software, which can make the development process slow. It can take a few seconds to complete one action, and then a few more seconds to do the next one.

It would be helpful to be able to connect in an easier way to data lakes or to structures that are residing in other cloud providers.

I have been using IBM Cloud Pak for Data for approximately 

The solution is stable.

I rate the scalability of IBM Cloud Pak for Data a nine out of ten.

I rate the scalability of IBM Cloud Pak for Data an eight out of ten.

The support can be slow to respond.

I rate the support from IBM Cloud Pak for Data a seven out of ten.

Neutral

The initial setup of IBM Cloud Pak for Data is easy because you can ask for it to deploy, but connecting the data sources can be a bit cumbersome sometimes.

The setup process requires one or two people, one with modeling expertise and the other with technical knowledge. If we were using the on-premise version, it would require more resources and personnel.

I rate the initial setup of IBM Cloud Pak for Data an eight out of ten.

IBM Cloud Pak for Data is expensive. If we include the training time and the machine learning, it's expensive. The cost of the execution is more reasonable.

We only need one or two people for maintaining the solution. This is thanks to the fact that we are using the cloud version of IBM Cloud Pak for Data. If we were using the on-premise version, it would require more resources and personnel.

My advice is if you are using Cloud Pak for Data on-premise, a highly technical team with knowledge of OpenShift and Red Hat operations and infrastructure setup is necessary. However, using the cloud service or SaaS version is much simpler and doesn't require as much technical expertise, making it a more favorable option.

It is a convenient platform that offers a variety of tools and algorithms, is easy to deploy and set up, and has many capabilities and functions that can be used in different projects. Additionally, it requires minimal coding compared to other vendor tools that I have used.

I rate IBM Cloud Pak for Data a nine out of ten.

I mainly use Cloud Pak in the FinTech and banking sectors.

Cloud Pak's most valuable features are IBM MQ, IBM App Connect, IBM API Connect, and ISPF.

Cloud Pak would be improved with integration with cloud service providers like Cloudera.

I've been working with Cloud Pak for four years.

Cloud Pak is 100% stable.

Cloud Pak uses cloud infrastructure, so it's easy to scale up and expand the services.

As partners, we get the best technical support because we get direct contact. Sometimes the tickets take a long time to get resolved, but in urgent cases, we get the response we expect.

The initial setup was a little complex as building the cloud infrastructure to work on-prem wasn't easy, so the culture for the confirmation from monolithic solution to microservices and cloud-enabled services wasn't easy to convince the customer with. The deployment took around two to three months.

Cloud Pak's cost is a little high.

I would 100% recommend Cloud Pak and would rate it eight out of ten.

We have used IBM Cloud Pak for Data for machine learning and data science use cases for various customers.

The most valuable features of IBM Cloud Pak for Data are the Watson Studio, where we can initiate more groups and write code. Additionally, Watson Machine Learning is available with many other services, such as APIs which you can plug the machine learning models.

Microsoft Azure has services, and you can have them be involved between any downstream applications for receiving results.

There is a solution that is part of IBM Cloud Pak for Data called Watson OpenScale. It is used to monitor the deployed models for the quality and fairness of the results. This is one area that needs a lot of improvement. 

The automated machine learning models can be created based on combining those two models on the IBM Cloud. This is an area that is needing improvement. 

AIOps learning is used to manage the different versions of models of the data using these same models that have been made. This needs to improve.

One large hurdle we face is the AI lifecycle, this is a feature I would like to see. IBM needs to simplify the installation process and the administration process, which should be more streamlined. There should be more focus on decoupling the individual solutions to allow customer flexibility. The customer might not have all of it, but now they can disable it. The lifecycle data provides an entire range starting from collection of data, organizing, analysis and improving it.

This happens in real life, where customers do not like to put all their eggs in the same basket. They need a diversified platform, so they may select IBM Cloud Pak for Data for the sorting processes, then for the machine learning, they can do it using Watson Machine Learning and Watson Studio, whereas maybe for the design innovation part, they may go for some other solution. This homogeneity, or the diversification, they should be able to achieve  All of their solutions should be made in a fashion that can be plug and play. The installation or the setup process should not be complex, and integration with other solutions should be available.  

I have been using IBM Cloud Pak for Data for approximately two years.

The initial setup was a little complex. It is not that user-friendly, and it needs quite a bit of expertise to install, the installation between various different vendors is even more difficult, such as deploying it on the IBM Cloud is relatively easier than having it installed in Amazon AWS or Microsoft Azure, or similar cloud service. This is an area where a lot of improvement is required. 

We have been working with IBM Cloud Pak for Data version 2.0. After two years there was 3.0, which was improved a lot. Many of the processes have been made simpler, but there is still some improvement needed. Now we are in version 4.0. 

IBM Cloud Pak for Data was not installed by us, it was installed by IBM.

IBM Cloud Pak for Data is a very useful tool because it has the entire gamut of tools. Starting from collecting the data from various sources using direct integration or through data virtualization and then organizing it into catalogs and applying their organization's policies or if they have other enforcement policies on top of that. We can build the model from that data because you can refine the data. There is a lot of AI inside this single solution, we could bring a lot of homogeneity into the organization. The data discovery process and the finding of the right set of data for the right problem will become much easier because it's a very good solution.

I rate IBM Cloud Pak for Data a seven out of ten.

The main problem that happens in machine learning projects is that people come in and then work on making the models, and then the model is deployed in production. However, I find that there is a bias in collecting the model, and the model has to be more advanced. The model which you have built, it's very much dependent on the data from which the model was built. 

Most of the solutions have features for this in the new version of the model, while they don't provide support for the version of the data or the parameters of the data on which the model was built. When a new person joins the organization, or if the engineer who has worked on it has left the organization, if a new person comes in, and they don't have that reference data on which the model was built then we have to start from scratch. This is an area where many of the solutions which are on the market don't have a solution. This is where IBM should focus more on providing a solution. It is a major area that not only IBM but other vendors, should start working on to provide a solution. I understand Microsoft is working on a solution for this, and they have a new concept that they are introducing which is called Machine Teaching. If there is only a partial resolution or solution for this problem to structure data, they have to focus more on bringing new, innovative solutions.