Overview

This CrewAI workflow creates a multi-agent research team that autonomously investigates topics, gathers sources, analyzes findings, and produces structured reports. Each agent has a specialized role — Researcher, Analyst, and Writer — and they collaborate sequentially to produce high-quality output.

Agent Roles

  • Research Agent: Searches the web, academic databases, and internal knowledge bases. Gathers raw data, citations, and source materials.
  • Analyst Agent: Processes raw research into structured insights. Identifies patterns, contradictions, and key takeaways.
  • Writer Agent: Transforms analysis into a polished report with executive summary, findings, and recommendations.

Workflow Steps

  1. Task Definition: User provides research topic, scope, and output format requirements.
  2. Parallel Research: Research Agent queries multiple sources simultaneously.
  3. Synthesis: Analyst Agent consolidates findings, removes duplicates, and identifies gaps.
  4. Report Generation: Writer Agent produces final output in requested format (Markdown, PDF, or DOCX).
  5. Review & Iterate: Optional human review step before final delivery.

Tools Required

  • CrewAI (Python framework)
  • OpenAI API or Anthropic API key
  • Optional: Serper API for web search, arXiv API for academic papers

Use Cases

  • Competitive intelligence reports
  • Market research and trend analysis
  • Academic literature reviews
  • Due diligence documentation