Concept

Understand the Knowledge Base concept in Future AGI. Store organizational content to ground synthetic data and provide reference context for evaluations.

A Knowledge Base is a store of your organization’s content — documents, policies, FAQs, SOPs — that Future AGI indexes and makes available across the platform.

Where it’s useful:

  • Synthetic data generation — Generated examples are grounded in your actual content, so the output reflects your domain, terminology, and procedures rather than generic text.
  • Evaluations — The KB supplies the reference context that evals check responses against. Instead of judging answers in a vacuum, the evaluator compares them to what your documents actually say — which is what makes hallucination detection meaningful.

What are the different knowledge assets?

  • Technical documentation and manuals
  • FAQs and troubleshooting guides
  • SOPs and process workflows
  • Training materials and HR policies
  • Legal documents and compliance information
  • Product descriptions and specifications

Features and Functionalities that are supported

  • Upload files in .doc.docx.pdf.txt, or .rtf (5MB max per file)
  • Drag-and-drop or bulk file upload supported
  • SDK support for large file ingestion
  • You can also remove files from a Knowledge Base or completely delete the Knowledge Base

Uploaded files are processed and categorised into:

StatusDescription
✅ SuccessfulContent extracted for knowledge base
🔄 ProcessingFile is being processed
❌ FailedUsers notified; file not usable

You can search for your knowledge bases using name and sort them based on the title and time updated.

Was this page helpful?

Questions & Discussion