HTTP APIs and AI
Since many AI systems are available via API (e.g., OpenAI's GPT-4), you can integrate their functionality into mappings in MapForce. To be able to use AI-powered data processing, you will need to create a Web service call to an API (e.g., OpenAI API, Azure OpenAI API, AWS AI Services, etc.).
Broad procedures
The broad procedures for configuring the AI functionality in MapForce are described below:
1.Insert a source component. This can be any data format supported by MapForce (e.g., an XML file, a database, a JSON file, etc.).
2.Insert a target component. A target component can be any data format supported by MapForce (e.g., an XML file, a database, a JSON file, etc.).
3.Insert a Web service call that will process source data. The processed source data will then be written to the target component.
4.Configure the Web service call by setting the API URL, your organization's secret API key and ID, JSON request and response structures. The JSON request and response structures can be auto-generated in MapForce or XMLSpy. The request and response structures will be visible in the Web service call component.
5.Continue configuring your mapping as required.
Use-case scenarios
The MapForceExamples folder includes the following examples of AI-powered mappings:
•SentimentAnalysis.mfd
•AI_ImageClassification.mfd
To be able to test these mappings, you will need your organization's login credentials. For more information about these examples and AI functionality in MapForce, see the following articles on the Altova website:
•AI-Based Support Request Sentiment Analysis Using MapForce and GPT-4
•AI-Based Database Image Classification with Altova MapForce