Welcome to the SmartPredict out-of-the-box AI use cases guide!

This guide will help take you from an absolute beginner or basic user to a confident, competent user able to run an end-to-end AI project with SmartPredict out-of-the-box AI use cases.

We have divided the guide into four articles:

- Good to know to get started

- Main steps of running an out-of-the-box AI use case

- Example 1: Running Credit Card Fraud Detection with the SmartPredict dataset (1st option)
- Example 2: Running the Credit Card Fraud detection with your dataset (2nd option)

Let's get started with "Good to know about SmartPredict out-of-the-box AI use cases".

It is a project that solves a real-world problem using AI. They are already created by SmartPredict's data scientists and domain experts team so that citizen data scientists - those who are not trained in data science - can successfully complete an end-to-end AI project in an easy and customizable way. It can be integrated with any platform to bring any business into the age of  AI to gain benefits. 

Currently, you can run the following ready-to-use AI use cases in SmartPredict:

Once you have successfully run a SmartPredict out-of-the-box AI use case, you can make the AI solution (described in the table above) operational in real-time in your enterprise’s platform or product through the generated API easily. Alternatively, you can test and use it in the SmartPredict AI platform itself. 

It is worth noting that SmartPredict's out-of-the-box AI use case is configurable and customizable, so AI models can be trained with business data to truly meet business needs to derive ROI. In technical terms, the generated ML workflow is also 100% customizable for more advanced personalization. 

• It’s free!
  You can run and integrate all AI use cases with your websites, applications, software, and all other platforms for free with the SmartPredict free plan to get started quickly.
• Really easy to use
  You will be able to perform all AI processes with a few clicks, in the easiest way in a matter of minutes.
• It’s collaborative
  The SmartPredict platform has implemented the “Team” concept so that all team members can see and work in real-time on the same AI project.
• A state of the art AI
  You will have the opportunity to use high-performance AI models that address business problems developed by our data science and domain experts. 
• Address real-world problem
  SmartPredit out-of-the-box AI cases are designed to meet the needs of businesses and give them accurate results that boost revenue, reduce costs and minimize risks. Most of these AI cases are already working in other companies and individuals to bring them into the AI era.

• Use case configuration
  Any out-of-the-box AI use case supports user configuration to refine business needs. It allows seamless real-world/business AI Project initialization. 
• ML workflow generation
  In most SmartPredict out-of-the-box AI use cases you run, SmartPredict generates two customizable flowcharts, which are: 
  - the ML workflow (resulted from the AI model training with your data), 
  - the deployment flowchart (that is fully deployed as a web service to put the trained model into production).
  All this gives you transparency and complete control over your AI project.
• Customizable
  Thanks to the generation of ML workflows as a 100% customizable flowchart, you can customize your out-of-the-box AI use case by coding with the notion of Custom modules. The sky's the limit! You can add any AI processing you like, such as data pre-processing, AI modeling, etc. With the generated deployment diagram, which is also 100% customizable and fully deployed as a web service, you can customize the AI project's production release by adding more functionality like calling APIs, etc.
• Versatile for any data enterprise
  AI use cases are conceived for business use. Thus, they are designed to receive data you use in your business in the most secure way possible. In addition, AI models can ingest your data for training, as AI use cases implement automated data pre-processing.
• Easy to integrate with external platforms
  Your AI project is automatically deployed as a web service, and an API is automatically generated to easily put it into production in applications, websites, and IT platforms.
• Easy for inferencing
  You can operationalize the deployed Machine Learning model right in the SmartPredict' s Predict space. That is to say, running live data points into the deployed machine learning algorithm (or “ML model”) or sending a request to the ML model to calculate an output such as indicating if a transaction is fraudulent or not, or generating a graph of the sales forecasting.

SmartPredict out-of-the-box AI use cases need to be run in the SmartPredict AI platform, so you should go to the platform. In the Explore Dashboard, you will find all the AI use cases that you can run right away, as you can see in the figure below. You only have to choose one by clicking on its card, and all the processes will run.