Ingest Content
Ingest content into Graphlit.
Last updated
Ingest content into Graphlit.
Last updated
Now that you've seen how to connect to the Graphlit API, you'll want to get started adding content to your project. Graphlit uses the term Ingestion for the process of uploading content into Graphlit, and starting the content processing workflow.
Graphlit provides end-to-end Knowledge ETL, which includes unstructured data ingestion, entity and text extraction, and vector embeddings.
Graphlit can ingest a wide variety of content sources into your project, as well as many different file types, such as Word documents, PDFs, Markdown text files, JPEG images, MP3 audio recordings and MP4 videos.
Graphlit automatically builds a knowledge graph from your content - both from the metadata embedded in files (such as title, author or keywords) and from the text extracted from documents or audio transcripts.
This knowledge graph maintains relationships between your content and people, places and things - which lets you do creative queries such as, which podcasts or Word documents mentioned "Machine Learning", or which Reddit posts and Slack messages discussed the company, OpenAI.
