The OpenAI_Connector is an OutSystems component that allows you to easily integrate with OpenAI’s Embedding API using the text-embedding-ada-002 model. It provides a streamlined way to generate vector embeddings from plain text, enabling the foundation for semantic search, document similarity, and RAG-based applications.

🔧 Installation & Setup

1. Install the Component
   Download and install the OpenAI_Connector component from the Forge into your OutSystems environment.

2. Create an OpenAI Account
   Go to https://platform.openai.com/account/api-keys and generate a secret API key.

3. Configure Site Properties

   • OpenAI_APIKey: Set your secret API key here.

   • OpenAI_EmbeddingModel: Defaults to text-embedding-ada-002 but can be changed to another supported model if needed.

   • These properties can be updated via Service Center > Factory > Site Properties.

4. Publish and Integrate
   After setting the required properties, publish the module and reference it in your application to start using the embedding functionality.

🧩 Available Public Methods

OpenAI_Embedding

• Input:

  • TextInput (Text): The text you want to convert into an embedding vector.

• Output:

  • EmbeddingList (List of Decimal): A list of float values representing the embedding.

This action sends the input text to OpenAI’s API and retrieves the corresponding vector representation.

✅ Use Cases

• Semantic document or FAQ search.

• Grouping similar content or clustering.

• RAG (Retrieval-Augmented Generation) applications.

• Text classification based on meaning rather than keywords.

📝 Notes

• The component does not store embeddings; it only retrieves them. You'll need to integrate it with a vector database (e.g., Pinecone, Qdrant, Weaviate) or your own storage solution for semantic operations.

• The component is built with simplicity and modularity in mind—perfect for POCs or production apps.

• No external dependencies beyond the OpenAI API.