RAG & Retrieval

Super-Rag with SAML


Super-Rag is performant Retrieval Augmented Generation pipeline specifcially created to be used by AI Agents. Super-Rag allows developers to ingest large amount of data and perform Q&A, Summarization and Computationl Q&A on that dataset.

Getting started

  1. Connect your a embedding database in the Superagent UI
  2. Create a workflow using SAML.
2 - superagent:
3 name: Rag Agent
4 llm: gpt-4-turbo-preview
5 prompt: Use the earning reports to answer any questions
6 superrag:
7 - index:
8 name: earnings
9 use_for: useful for answering questions about earning reports.
10 urls:
11 - https://digitalassets.tesla.com/tesla-contents/image/upload/IR/TSLA-Q4-2023-Update.pdf
12 - https://s2.q4cdn.com/299287126/files/doc_financials/2023/q4/AMZN-Q4-2023-Earnings-Release.pdf
  1. Query the workflow using the API or SDKs

Running Super-Rag as a stand-alone service

Super-Rag comes as a stand-alone REST API. You can read more on how to setup here. Alternatively you can run the free Cloud API to ingest/retrieve data.