phase3_agents.py

import os
from phi.agent import Agent
from phi.model.openai import OpenAIChat
import openai
import markdown2
import pdfkit
from datetime import datetime
from bni_agent import get_bni_benefits
from rag_agent import recommend_bni_connections

# Load environment variables (API keys, etc.)
from dotenv import load_dotenv
load_dotenv()

#####################################################################################
#                                    PHASE 3                                        #
#####################################################################################

##############################
# 1️⃣ Reasoning Agent        #
##############################
reasoning_agent = Agent(
    name="Reasoning Agent",
    model=OpenAIChat(id="gpt-4o"),
    description="Processes all collected data and generates structured AI adoption strategies.",
    show_tool_calls=True,
    markdown=True,
)

def generate_ai_strategy(company_data: str, industry_trends: str) -> str:
    """Generates a structured AI adoption strategy based on company details and industry insights."""
    query = f"""
    You are an AI business strategist analyzing a company's AI adoption potential. Given:
    - **Company Overview:** {company_data}
    - **Industry Trends:** {industry_trends}
    
    Generate a structured AI strategy including:
    1. Key AI Opportunities
    """
    response = reasoning_agent.run(query)
    return response.content  # Returns the generated AI strategy


##############################
# 2️⃣ AI Integration Advisor  #
##############################
# ai_integration_agent = Agent(
#     name="AI Integration Advisor",
#     model=OpenAIChat(id="gpt-4o"),
#     description="Suggests AI implementation strategies based on industry insights and company operations.",
#     show_tool_calls=True,
#     markdown=True,
# )

# def suggest_ai_integration(company_data: str, ai_strategy: str) -> str:
#     """Suggests a structured AI implementation plan."""
#     query = f"""
#     Based on the AI adoption strategy:
#     - **Company Context:** {company_data}
#     - **AI Strategy Summary:** {ai_strategy}

#     Provide a structured AI implementation plan including:
#     1. Step-by-step AI adoption process
#     2. Required AI Technologies & Infrastructure
#     3. Workforce training & AI skill development
#     4. Risk considerations (data security, compliance, ethical AI)
#     5. Key performance indicators (KPIs) for AI success.
#     """
#     response = ai_integration_agent.run(query)
#     return response.content  # Returns AI integration plan


##############################
# 3️⃣ Revenue Growth Agent    #
##############################
revenue_growth_agent = Agent(
    name="Revenue Growth Agent",
    model=OpenAIChat(id="gpt-4o"),
    description="Identifies AI-driven opportunities to enhance revenue and efficiency.",
    show_tool_calls=True,
    markdown=True,
)

def identify_revenue_opportunities(company_data: str, ai_strategy: str) -> str:
    """Identifies AI-driven revenue generation opportunities for a business."""
    query = f"""
    You are an AI business analyst identifying practical AI-driven revenue growth opportunities for:
    
    Company Overview: {company_data}
    AI Strategy: {ai_strategy}
    Provide a concise, single-paragraph summary covering:
    
    AI-based Monetization Strategies – How can the company create new revenue streams using AI?
    Cost Reduction & Operational Efficiency – How can AI streamline operations and reduce costs?
    Market Expansion – How can AI open new markets or customer segments?
    Competitive Positioning – How can AI help differentiate the company from competitors?
    Keep it short, crisp, and actionable, avoiding complex or unnecessary details.
    """
    response = revenue_growth_agent.run(query)
    return response.content  # Returns revenue opportunities


##############################
# 4️⃣ BNI Agent: Membership Benefits #
##############################
def get_bni_membership_benefits(company_data: str) -> str:
    """Fetches BNI membership benefits tailored to the user's company."""
    return get_bni_benefits(company_data)


##############################
# 5️⃣ BNI RAG Agent: Pearl Chapter Connections #
##############################
def get_bni_recommendations(company_data: str) -> str:
    """Recommends relevant BNI Pearl Chapter members based on the user's company data."""
    return recommend_bni_connections(company_data)


##############################
# 6️⃣ Report Generation Agent #
##############################
# Ensure your OpenAI API key is set in your environment variables
openai.api_key = os.getenv("OPENAI_API_KEY")

def generate_report(company_name: str, company_data: str,
                    ai_strategy: str,
                    revenue_opportunities: str,
                    bni_benefits: str,
                    bni_connections: str) -> str:
    """
    Generates a structured AI strategy report in PDF format using `pdfkit`.
    """

    # 📝 **Step 1: Prepare the Report Content**
    report_content = f"""
    # AI Strategy Report for {company_name}
    
    ## 1️⃣ Company Overview
    {company_data}
    
    ## 2️⃣ How BNI Can Help Your Business
    {bni_benefits}

    ## 3️⃣ Recommended BNI Pearl Chapter Connections
    {bni_connections}

    ## 4️⃣ AI Adoption Strategy
    {ai_strategy}
    
    ## 5️⃣ Revenue Growth Opportunities
    {revenue_opportunities}
    """

    # 🎯 **Step 2: Use OpenAI GPT-4o-mini to Format Markdown**
    response = openai.chat.completions.create(
        model="gpt-4o-mini",
        messages=[
        {"role": "system", "content": "You are an expert in formatting business reports in Markdown."},
        {"role": "user", "content": f"Format the following report in well-structured Markdown. Use bold headings, bullet points, and clear sections. Ensure the text is concise, professional, and structured for readability. Do not include any emojis or unnecessary text:\n\n{report_content}"}
        ],
        max_tokens=5000,
        temperature=0.5,
    )

    enhanced_markdown = response.choices[0].message.content  # Extract the formatted response

    # 🖼 **Step 3: Convert Enhanced Markdown to HTML**
    html_report = markdown2.markdown(enhanced_markdown)

    # 📄 **Step 4: Define the HTML Template with Styling**
    html_template = f"""
    <html>
    <head>
        <style>
            body {{
                font-family: 'Times New Roman', serif;
                margin: 50px;
                position: relative;
            }}
            .watermark {{
                position: fixed;
                top: 50%;
                left: 50%;
                transform: translate(-50%, -50%);
                font-size: 100px;
                color: rgba(200, 200, 200, 0.2);
                z-index: -1;
                user-select: none;
            }}
            .outline {{
                border: 2px solid #000;
                padding: 20px;
                margin-bottom: 20px;
            }}
            h1, h2, h3, h4, h5, h6 {{
                color: #333;
            }}
            p {{
                font-size: 14pt;
                line-height: 1.6;
                color: #000;
            }}
            .footer {{
                position: fixed;
                bottom: 20px;
                left: 0;
                width: 100%;
                text-align: center;
                font-size: 12pt;
                color: #555;
                border-top: 1px solid #ccc;
                padding-top: 10px;
            }}
        </style>
    </head>
    <body>
        <div class="watermark">Giant Analytics</div>
        <div class="outline">
            {html_report}
        </div>
        <div class="footer">
            🚀 Powered by <strong>Giant Analytics</strong> | Contact Us: (+91) 96533 20500
        </div>
    </body>
    </html>
    """


    # 📌 **Step 5: Generate the PDF**
    current_date = datetime.now().strftime("%Y-%m-%d")
    pdf_filename = f"{company_name}_AI_Report_{current_date}.pdf"

    # Path to wkhtmltopdf executable (Make sure it's installed)
    path_to_wkhtmltopdf = '/usr/bin/wkhtmltopdf'  # Update this path if different
    config = pdfkit.configuration(wkhtmltopdf=path_to_wkhtmltopdf)

    # Convert HTML to PDF
    pdfkit.from_string(html_template, pdf_filename, configuration=config)

    return pdf_filename  # Returns the generated PDF filename