Hi everyone,
Ive made a form on ontology that accepts user data and an attachment, I wanted to know of a way I could make a pipeline that helps me run doc extraction on those files, I’ve been having trouble in accessing them as Code repo pipelines dont let me import ontology data, how must I proceed?
To access ontology data in your pipelines, you can create a materialization of your ontology object type. This will give you the most up to date version of your data (raw data & user edits).
Hi, thank you for your response. I’ve created a form using Ontology where users can submit data along with file attachments. These uploaded documents are stored as RIDs in an Ontology table.
Now, I want to perform document extraction on these attachments using a vision-language model. However, I’m facing a challenge: I’m not sure how to access these attachments from within the Pipeline Builder or CodeRepo Transforms. Both of these require explicit input and output datasets and don’t seem to support working with files stored as RIDs or creating a directory-like structure.
Could you please guide me on how to access and process these attached documents for extraction within the Pipeline framework?