Showcases

Real-world applications built with LangGraphGo.

DeerFlow Screenshot
Deep Research

DeerFlow

DeerFlow is a deep research system based on multi-agent architecture (Go port of ByteDance DeerFlow). It orchestrates Planner, Researcher, and Reporter agents to conduct deep research on a given topic and generate detailed reports.

Features: Modern Web Interface, Real-time SSE status updates, CLI support.

go build -o deerflow ./showcases/deerflow ./deerflow
Multi-Agent Collaboration

Open Deep Research

Go implementation of langchain-ai/open_deep_research. This is a hierarchical multi-agent system where a Supervisor agent coordinates multiple parallel Researcher agents to conduct comprehensive research on complex questions.

Features: Parallel execution, Tavily search integration, research compression, configurable models.

cd showcases/open_deep_research export OPENAI_API_KEY="..." export TAVILY_API_KEY="..." go run *.go "Your research question"
Tool Usage

DeepAgents

An example agent with file system access and sub-agent delegation capabilities. Demonstrates how to let agents use tools for file reading/writing, managing todo lists, and handling complex tasks.

cd showcases/deepagents export OPENAI_API_KEY="..." go run main.go
Financial AI

Trading Agents

A multi-agent LLM-driven financial trading framework that simulates a professional trading firm. Features specialized agents including fundamental analysts, sentiment analysts, news analysts, technical analysts, research teams, traders, and risk management teams collaborating to analyze stocks and make informed trading decisions.

Features: Real-time market data integration, Backend API server, CLI interface, Web dashboard, Comprehensive logging and tracing.

cd showcases/trading_agents export OPENAI_API_KEY="..." export ALPHA_VANTAGE_API_KEY="..." go build -o bin/trading-cli ./cli ./bin/trading-cli -cmd analyze -symbol AAPL