┌─────────────────────────────────────────────────────────────────────────────┐
│                         MULTI-AGENT SYSTEM                                 │
├─────────────────────────────────────────────────────────────────────────────┤
│                                                                             │
│  ┌─────────────────────────────────────────────────────────────────────┐    │
│  │                        ORCHESTRATOR                                 │    │
│  │  (Manages agents, coordinates tasks, aggregates results)           │    │
│  └─────────────────────────────────────────────────────────────────────┘    │
│                           │           │           │                        │
│                           ▼           ▼           ▼                        │
│  ┌─────────────────┐ ┌─────────────────┐ ┌─────────────────┐             │
│  │   DATA AGENT    │ │  RESEARCH AGENT │ │  SENTIMENT AGENT│             │
│  │  (Collects data)│ │  (Analyzes data)│ │ (News/ Social)  │             │
│  └─────────────────┘ └─────────────────┘ └─────────────────┘             │
│                           │           │           │                        │
│                           ▼           ▼           ▼                        │
│  ┌─────────────────────────────────────────────────────────────────────┐    │
│  │                        SHARED MEMORY                               │    │
│  │  (Vector database, knowledge graph, message bus)                   │    │
│  └─────────────────────────────────────────────────────────────────────┘    │
│                           │                                                │
│                           ▼                                                │
│  ┌─────────────────────────────────────────────────────────────────────┐    │
│  │                        EXECUTION AGENT                             │    │
│  │  (Executes trades, generates reports)                              │    │
│  └─────────────────────────────────────────────────────────────────────┘    │
│                                                                             │
└─────────────────────────────────────────────────────────────────────────────┘

from autogen import AssistantAgent, UserProxyAgent, GroupChat, GroupChatManager

# Define agents
research_agent = AssistantAgent(
    name="ResearchAgent",
    system_message="You are a financial research analyst. Provide insights on companies."
)

sentiment_agent = AssistantAgent(
    name="SentimentAgent",
    system_message="You are a sentiment analyst. Analyze market sentiment."
)

technical_agent = AssistantAgent(
    name="TechnicalAgent",
    system_message="You are a technical analyst. Analyze technical indicators."
)

portfolio_agent = AssistantAgent(
    name="PortfolioAgent",
    system_message="You are a portfolio manager. Make allocation decisions."
)

# Define group chat
group_chat = GroupChat(
    agents=[research_agent, sentiment_agent, technical_agent, portfolio_agent],
    messages=[],
    max_round=20,
    speaker_selection_method="round_robin"
)

manager = GroupChatManager(groupchat=group_chat)

# Start the conversation
user_proxy = UserProxyAgent(name="UserProxy", human_input_mode="NEVER")
user_proxy.initiate_chat(
    manager,
    message="Analyze the tech sector and suggest a portfolio allocation."
)

from crewai import Agent, Task, Crew

research_agent = Agent(
    role="Research Analyst",
    goal="Conduct thorough research on companies",
    backstory="You are a highly experienced financial analyst."
)

sentiment_agent = Agent(
    role="Sentiment Analyst",
    goal="Analyze market sentiment from news and social media",
    backstory="You are an expert in NLP and sentiment analysis."
)

task1 = Task(
    description="Research Company X and provide a summary.",
    agent=research_agent
)

task2 = Task(
    description="Analyze sentiment around Company X.",
    agent=sentiment_agent
)

crew = Crew(
    agents=[research_agent, sentiment_agent],
    tasks=[task1, task2]
)

result = crew.kickoff()

from langgraph.graph import StateGraph

# Define state
class AgentState:
    def __init__(self):
        self.messages = []
        self.data = {}

# Define agent functions
def research_agent(state):
    # Analyze and update state
    state.messages.append("Research conducted.")
    return state

def decision_agent(state):
    # Decide based on state
    if state.data.get("risk") < 0.5:
        state.messages.append("Decision: Buy")
    else:
        state.messages.append("Decision: Hold")
    return state

# Build graph
graph = StateGraph(AgentState)
graph.add_node("research", research_agent)
graph.add_node("decision", decision_agent)
graph.set_entry_point("research")
graph.add_edge("research", "decision")
graph.add_edge("decision", END)

Data Agent → Research Agent → Portfolio Agent → Execution Agent

        ┌─────────────────┐
        │   Data Agent    │
        └────────┬────────┘
                 │
        ┌────────┼────────┐
        │        │        │
        ▼        ▼        ▼
  ┌─────────┐┌─────────┐┌─────────┐
  │Research ││Sentiment││Technical│
  │ Agent   ││ Agent   ││ Agent   │
  └─────────┘└─────────┘└─────────┘
        │        │        │
        └────────┼────────┘
                 ▼
        ┌─────────────────┐
        │  Aggregator     │
        └─────────────────┘

        ┌─────────────────────────────────────┐
        │          MODERATOR                  │
        └─────────────────────────────────────┘
                 │
        ┌────────┼────────┐
        │        │        │
        ▼        ▼        ▼
  ┌─────────┐┌─────────┐┌─────────┐
  │  Bull   ││  Bear   ││ Neutral │
  │  Agent  ││  Agent  ││ Agent   │
  └─────────┘└─────────┘└─────────┘
        │        │        │
        └────────┼────────┘
                 ▼
        ┌─────────────────┐
        │  Decision Agent │
        └─────────────────┘

        ┌─────────────────────────────────────┐
        │          MANAGER AGENT              │
        └─────────────────────────────────────┘
                 │
        ┌────────┼────────┐
        │        │        │
        ▼        ▼        ▼
  ┌─────────┐┌─────────┐┌─────────┐
  │Sub-agent││Sub-agent││Sub-agent│
  │   1     ││   2     ││   3     │
  └─────────┘└─────────┘└─────────┘

┌────────────┐    ┌────────────┐    ┌────────────┐
│    Data    │───▶│   Message  │───▶│  Research  │
│    Agent   │    │    Bus     │    │   Agent    │
└────────────┘    └────────────┘    └────────────┘
                       │
                       ▼
                 ┌────────────┐
                 │  Portfolio │
                 │   Agent    │
                 └────────────┘

1. Data Agent: Collects market data, news, and financials.
2. Bull Agent: Presents the bullish case.
3. Bear Agent: Presents the bearish case.
4. Value Agent: Presents the value case.
5. Technical Agent: Presents the technical analysis.
6. Risk Agent: Evaluates the risk of each case.
7. Moderator: Summarizes the discussion.
8. Decision Agent: Makes the final decision.

You are an optimistic investment analyst. Your role is to present the bullish case for the proposed investment.

Key points to highlight:
1. Growth potential
2. Competitive advantages
3. Positive catalysts
4. Favorable market trends

Be persuasive but evidence-based.

You are a cautious investment analyst. Your role is to present the bearish case for the proposed investment.

Key points to highlight:
1. Risks and challenges
2. Competitive threats
3. Negative catalysts
4. Unfavorable market trends

Be thorough but fair.

You are the decision-maker for the investment committee.

Based on the arguments presented:
1. What is your decision? (Buy/Hold/Sell)
2. What is the conviction level? (High/Medium/Low)
3. What are the key reasons for your decision?
4. What is the recommended position size?

Synthesize all the arguments and make a reasoned decision.