How we built our multi-agent research system
AI Summary
Anthropic's engineering writeup on building Claude's Research feature, an orchestrator-worker multi-agent system where a LeadResearcher spawns parallel subagents that each search independently before a CitationAgent adds sourcing. It reports a 90.2% improvement over single-agent Opus 4 on internal evals, explains that token usage and parallel tool calls are the dominant performance levers, and details prompt-engineering and production-reliability lessons (rainbow deployments, LLM-as-judge evals, human testing) learned moving from prototype to production.
Original excerpt
Our Research feature uses multiple Claude agents to explore complex topics more effectively. We share the engineering challenges and the lessons we learned from building this system.
Claude now has Research capabilities that allow it to search across the web, Google Workspace, and any integrations to accomplish complex tasks.
The journey of this multi-agent system from prototype to production taught us critical lessons about system architecture, tool design, and prompt engineering. A multi-agent system consists of multiple agents (LLMs autonomously using tools in a loop) working together. Our Research feature involves an agent that plans a research process based on user queries, and then uses…
Frequently asked questions
What is "How we built our multi-agent research system" about?
Anthropic's engineering writeup on building Claude's Research feature, an orchestrator-worker multi-agent system where a LeadResearcher spawns parallel subagents that each search independently before a CitationAgent adds sourcing. It reports a 90.2% improvement over single-agent Opus 4 on internal e…
Who wrote "How we built our multi-agent research system"?
"How we built our multi-agent research system" was written by Jeremy Hadfield, Barry Zhang, Kenneth Lien, Florian Scholz, Jeremy Fox, Daniel Ford. It is curated in the AI Deep Research vault on Burn 451, which covers ai research agents & methodology.
How can I read more content from AI Deep Research?
The complete AI Deep Research reading list is available at burn451.cloud/vault/deep-research. Each article includes an AI-generated summary so you can decide what to read in seconds. Connect the Burn 451 MCP server to Claude or Cursor to query all AI Deep Research articles as live AI context.
Can I use "How we built our multi-agent research system" with Claude or Cursor?
Yes. Install the burn-mcp-server npm package and connect it to Claude Desktop, Claude Code, or Cursor. Once connected, your AI can search and reference this article and the full AI Deep Research vault in real time — no manual copy-paste required.
11 more articles in this vault.
Import the full AI Deep Research vault to Burn 451 and build your own knowledge base.
Content attributed to the original author (Jeremy Hadfield, Barry Zhang, Kenneth Lien, Florian Scholz, Jeremy Fox, Daniel Ford). Burn 451 curates publicly available writing as a reading index. For removal requests, contact @hawking520.