Back to Projects
Multi-Agent System

Research Brief Orchestrator

Multi-agent research pipeline that produces citation-grounded research briefs with real-time streaming.

Tech Stack

PythonLangGraphFastAPIReactRAGLangChainTypeScriptWebSockets

Overview

The Research Brief Orchestrator is a production-grade multi-agent system built with LangGraph, FastAPI, and React. Given a research topic, it orchestrates a pipeline of specialized agents: a planner that breaks down the query, a retrieval agent that fetches and ranks sources, a verification agent that filters unreliable citations, and a QA agent that checks factual consistency before producing the final brief. The system streams progress updates to the frontend in real time.

Key Highlights

  • LangGraph-based multi-agent orchestration with planning, retrieval, verification, and QA stages

  • Citation-grounded outputs — every claim is traceable to a verified source

  • Real-time streaming via WebSockets with per-stage progress indicators

  • Modular agent design allows swapping retrieval backends (web, vector DB, academic APIs)

  • FastAPI backend with async processing for concurrent research pipelines