Skip to main content
This guide will help you get started with Captide quickly, showing you how to set up both the REST API and the Document Auditing JavaScript SDK to embed corporate disclosure intelligence into your systems.
Before you begin, make sure you request a Captide test API key.

Using the REST API

Authentication

All API requests require authentication using your API key in the header:

Example 1: Retrieve Company Information

Use the /companies/list endpoint when you want to search companies in Captide and retrieve their company id to use as parameter in other endpoints.
Use the response to retrieve company information and use the id to retrieve company docs:

Example 2: Retrieve Company Documents

Use the /documents/list endpoint to search for company documents by applying filters such as documentCategories, formTypes, or by specifying a company’s tickers or id. This allows you to retrieve the metadata of the latest corporate disclosures of publicly listed companies.
In the example below, we use the id of Santos to retrieve its latest interim-reports:
The response below contains metadata for each document. The document id is particularly important, as it is used to retrieve the document file and its formatted Markdown version.
Once target documents are identified, we can use their id with the /documents/ endpoint to retrieve the URLs for the document files.
The response includes two additional fields: originalFileUrl and markdownFileUrl. These provide temporary access to the original document file and its formatted Markdown version, which is optimized for downstream analysis and LLM ingestion.

Example 3: Prompt-Based Chunk Retrieval

With Captide, you can create RAG-based applications and tools without the need to collect, preprocess, chunk, or store document excerpts yourself. Simply use the /rag/chunk-retrieval endpoint to send natural language requests, and you’ll receive relevant document excerpts from our database containing all the information needed to fulfill your query.
The response includes a list of relevant document chunks from the Captide database, along with their associated metadata.
You can then use the relevant chunks as context for AI-generated responses, and even combine them with context from other sources—such as internal data or structured databases.

Example 4: AI Agent Response

Captide enables you to get answers to any question grounded in information from corporate disclosures with full traceability. This allows you to instantly extract insights and metrics from millions of financial filings, or to build multi-agent systems that accelerate and enhance investment research. Below is an example query submitted to Captide’s AI agent.
The response includes the final answer in the content field, along with source identifiers (formatted as #abcd1234) embedded within the text. These identifiers link to the sourceMap, allowing each part of the answer to be traced back to the original document excerpts used.
The sourceMap can be integrated with Captide’s Document Auditing SDK to render relevant content chunks in frontend applications, enabling accurate source auditing.

Example 5: Streaming AI Agent Response

The /rag/agent-response-stream endpoint streams Captide’s AI agent responses in real time using Server-Sent Events (SSE). It’s useful in chat interfaces where you want to show the agent’s reasoning steps and display the response as it’s being generated, rather than waiting for the full message.
The SSE stream response contains the following event types:
  • thinking events – Indicate the agent’s internal progress and reasoning stages (e.g., searching, analyzing).
  • message/partial events – Stream the response incrementally, sending words or phrases as they’re generated.
  • metadata event – Provides associated document metadata, source links, and citations.
All message/partial events can be combined to reconstruct the full response and, as with the /rag/agent-response endpoint, the final metadata event can be used with Captide’s Document Auditing SDK to display relevant content chunks in frontend applications, enabling accurate source traceability and auditing.

Using the Document Auditing JavaScript SDK

Installation

Install the Captide SDK in your project:

Integration Example

Here’s how to integrate the document viewer into your React application:
For a live example of Captide in action, visit app.captide.co/chat.

Next Steps