I want to train my own data (folder containing PDF files) and chat with those data only. is it possible to do RAG using this Azure Openai connector?
Hello @JMATHEW.
To achieve this you need to set up a solution to implement Retrieved Augmented Generation.
In this example, were you have OpenAI, you can have this component Azure OpenAI Connector.
It is not a complex solution, but it is long explanation, can you please check this session we did explain how to implement it?
Let me know if this helps, feel free to make follow up questions, depending on your use case.
- Bruno
Hi Bruno,
Is there any chance you could provide the .oml file so that we could follow through and learn?
I am new but would like to try Open AI with OutSystems.
Thank you!
Howard
I wrote a two article series on how to implement a simple rag pipeline using OpenAI. Including a demo application on Forge. I hope you will find the articles helpful to build your own.,
Part 1 https://itnext.io/outsystems-with-openai-embeddings-and-qdrant-vector-database-find-similar-fbd9a5e4760a
Part 2 https://itnext.io/outsystems-openai-embeddings-and-qdrant-vector-database-answer-right-5b511764e6fa
However it does not feature content extraction from documents. For this you would have to add another extraction step. e.g. using AWS Textract Forge component for content extraction.