Release Notes

EDGAR Knowledge Source Enhancements

We are enhancing Harvey's EDGAR to be able to perform agentic, iterative search over all SEC filing types to extract relevant insights.

Release Date
Jan 14, 2026
Categories
Knowledge
Release Type
Regional Availability
US, EU, AU

What’s New

We’ve enhanced Harvey’s EDGAR Knowledge Source to make it smarter, broader, and more useful. The new version includes:

  • Improved retrieval quality for more relevant search results
  • Support for all SEC filing types, not just a limited subset
  • Better handling of broad, exploratory queries (i.e. open-ended prompts)
EDGAR Screenshot

Why It Matters

These enhancements will help you uncover relevant insights from filings more efficiently and answer open-ended questions directly within Harvey.

How to Use

  1. Open Sources.
  2. Select EDGAR from the list of options.
  3. Filter and select specific filing types (if known) for improved accuracy. Even if you select all filing types in the modal, Harvey will still prioritize the filing types you explicitly reference in your query.
    1. Alternatively, type @ in your prompt to mention specific filing types for Harvey to prioritize.
  4. Write your prompt in natural language. Be as specific as possible about:
    1. The company or industry
    2. The filing types (if relevant)
    3. The sections you care about (e.g., exhibits, risk factors, MD&A, executive compensation)
  5. Click Ask Harvey. Harvey will retrieve and synthesize information from the relevant filings and return an answer with citations.

Notes and Limitations

  • For customers with existing EDGAR access: If customers already have EDGAR enabled, these enhancements will automatically roll out.
  • We currently show sources, but we don’t yet highlight the exact text within those sources that supports an answer. We are working on adding this ability in future.

FAQs

Q: How will this help my workflow?

The improved EDGAR will make it easier to locate, summarize, and analyze information across filings—reducing manual research time and improving the accuracy of insights generated within Harvey.