Skip to main content

​
Research Agent Engine Overview

The Research Agent Engine leverages AI-powered search providers to conduct comprehensive web research. It can search the web, synthesize information from multiple sources, and return structured insights with source citations.

​
Engine Inputs

The Research Agent Engine Configuration has the following parameters:
  • prompt: required. The research query or question to investigate.
  • provider: optional. The search provider to use (default: perplexity). Options:
    • perplexity: Perplexity AI search (general web, conversational)
    • exa: Exa AI-powered semantic search (structured/academic)
  • use_deep_search: optional. Enable deep search for more comprehensive results (default: False). Higher cost but more thorough.
  • include_sources: optional. Include source URLs in the response (default: True).
  • output_schema: optional. JSON schema for structured response format. For Exa, providing a schema routes to the Research API; for Perplexity, schemas work with standard or deep search.
See Template Strings for dynamic parameter configuration.

​
Engine Output

The output includes researched information with optional source citations. When an output schema is defined, the response will match that structure.

​
Example Usage

1

Create an Agent

Click on the “Add Agent” button in the top right corner of the Agents page.
Enter a name and an optional description of your Agent.
2

Select the Research Agent Engine

3

Configure the engine

$ starts a template string
  • prompt: $prompt
  • provider: perplexity
  • use_deep_search: False
  • include_sources: True
  • output_schema: Copy and paste the JSON schema below:
{
  "type": "object",
  "properties": {
    "title": {
      "type": "string",
      "description": "Main title or heading"
    },
    "summary": {
      "type": "string",
      "description": "Brief summary of the research findings"
    },
    "key_points": {
      "type": "array",
      "items": { "type": "string" },
      "description": "List of key points from the research"
    }
  },
  "required": ["title", "summary"]
}
4

Create the Agent

Hit the Create button.
5

Run a job

Create a new job and provide a research prompt like:
  • “What are the latest developments in AI-powered document processing?”
  • “Research the competitive landscape for enterprise data warehouses”

​
Search Providers

ProviderDescriptionBest For
PerplexityConversational AI search (default)General research, complex questions
ExaAI-powered semantic searchStructured/academic research, technical topics

​
Use Cases

  • Market research: Gather competitive intelligence and market trends
  • Due diligence: Research companies, products, or individuals
  • Content creation: Gather source material for articles and reports
  • Fact-checking: Verify claims with web sources
  • Technical research: Find documentation, tutorials, and best practices