What is our primary use case?
In our team, we construct different statistical models to resolve things for clients. We do modelling and segmentation to determine a customer's lifetime value. We do deep learning and protective analysis.
What is most valuable?
The processing of data is good. We can use many different types of data, including images and videos. We can use different libraries to do better preprocessing of different types of data. We can do different models in recommendation systems for things like videos and sales strategy. We can do language processing or sentiment analysis to predict things for our clients.
We can design and develop machine learning applications that can predict events. We can use these on libraries to solve complex problems when we have a lot of data.
It's possible to use the solution with other tools. It's very agile.
In the financial advisory and portfolio management space, several budget management applications are now available in the market. These have machine learning based functionality. In Python, I use different machine learning algorithms to enable customers to keep track of their expenses and provide recommendations on better savings. These are machine learning algorithms that customize financial portfolios by looking at income rate tolerance and preferences, et cetera.
What needs improvement?
I've worked with file detection for secure transactions. I use a machine learning model to predict events related to security transactions by predicting possible routes in advance. They need to improve the part of deep learning that deals with security and the prediction of fraud.
It can be expensive to deploy a lot of models in the cloud, especially if there are large amounts of data.
What do I think about the stability of the solution?
I haven't had issues with stability.
What do I think about the scalability of the solution?
It is possible to scale the solution, however, it depends on the depth and quality of the data. If you have bad data or low-quality data, you need to fix the process in order to get better data. However, it is a very good tool for constructing a lot of algorithms and combining different tools.
How are customer service and support?
On the web and through the portal, there are a lot of different answers to problems that you can look at. Customer service has been fine overall.
How would you rate customer service and support?
How was the initial setup?
The initial setup may be straightforward or complex, depending on the behavior of the data. Sometimes it is easy to process the data as it is of better quality. However, if the data isn't as good, it can be more complex.
What was our ROI?
We have seen a 40% ROI. I've used it to make strategic actions for clients that have helped them uncover relevant insights that have led to cost savings. With this tool, they can explore decisions and make better overall decision-making.
What other advice do I have?
We're just a customer.
It's a good tool. It's easy. I can use it in many different ways and for many different use cases.
I'd recommend the product for data science use cases. It is a robust tool and very useful for constructing historical models. It has artificial intelligence built in.
I'd rate the product eight out of ten.
