Choose a research task

Start with your question, then load the guide and API reference needed for that step. All the research operations below read existing data. They do not acquire missing documents, start extraction, or submit a filing.

Your task	Start here	Expected result
Find filings by state, company, line, date, or native reference	Find the right filings	Filing IDs and metadata; a continuation when more results are available
Find policy wording or an exclusion	Search document text	Candidate filings and matching file IDs, with passages when available
Research a carrier, group, or MGA	Research a carrier or MGA	Directory identities connected to relevant filings
Read a document, resume later, or cite evidence	Read and cite document evidence	Revision-bound text chunks or a pinned original download
Inspect forms or understand a filed rate change	Interpret forms and rate impact	Stored extracted results with explicit coverage limits
Continue a list or recover a failed request	Pagination and recovery	A safe next request without mixing queries or text revisions

Connect once

Follow Quickstart, provide an ordinary user API key through EFFECTIVE_API_KEY, and verify the returned user and tenant. Filing research also requires permission to view SERFF filings; content reads enforce file access and export policy. Set API_BASE to the host for your key, without /api/v2 or a trailing slash:

export API_BASE='https://canary.effectiveai.app'
curl --fail-with-body --silent --show-error \
  "$API_BASE/api/v2/users/me" \
  --header "Authorization: Bearer $EFFECTIVE_API_KEY"

The guides show short HTTP requests relative to this host. Send the same Authorization: Bearer header on every API request and use Content-Type: application/json for JSON bodies. Replace placeholders such as {fileId} with returned IDs and URL-encode query values, including cursors. Response fragments are abbreviated; opaque values shown in angle brackets are illustrative. Keep your key out of saved request logs and send it only to your API host.

Load more context when needed

The guides explain workflows. The generated API reference owns request fields, enums, response schemas, and operation-specific limits. Use the documentation index to find a page, append .md for readable Markdown, or use the API catalog to retrieve OpenAPI. For every research result, keep its source reference, returned IDs, filters and observation time. State whether you read metadata, extracted text, or original bytes. A complete traversal of available records does not establish complete coverage of a regulator’s filings.

Getting Started

Research guides

Connect once

Load more context when needed

​Connect once

​Load more context when needed

Connect once

Load more context when needed