feat: loan underwriting agent added (#1344)

python/examples/advanced_agents/loan_underwriter_agent/.env.example

@@ -0,0 +1,2 @@
+GEMINI_API_KEY=KEY
+COMPOSIO_API_KEY=KEY

python/examples/advanced_agents/loan_underwriter_agent/main.py

@@ -0,0 +1,128 @@
+import streamlit as st
+import os
+from google import genai
+from composio_gemini import ComposioToolSet, App, Action
+from dotenv import load_dotenv
+from google.genai import types
+
+def initialize_ai():
+    load_dotenv()
+    toolset = ComposioToolSet()
+    tools = toolset.get_tools(actions=[
+        Action.GOOGLEDOCS_CREATE_DOCUMENT_MARKDOWN
+    ])
+
+    config = types.GenerateContentConfig(tools=tools) # type: ignore
+    client = genai.Client(api_key=os.getenv('GEMINI_API_KEY'))
+    return client, config
+
+def evaluate_loan_application(data, client, config):
+    prompt = f"""
+    Analyze this business loan application:
+    Business Name: {data['business_name']}
+    Industry: {data['industry']}
+    Years in Operation: {data['years_in_operation']}
+    Annual Revenue: ${data['annual_revenue']:,}
+    Credit Score: {data['credit_score']}
+    Loan Amount Requested: ${data['loan_amount']:,}
+    Current Outstanding Debt: ${data['current_debt']:,}
+    
+    Provide a detailed evaluation including:
+    1. Loan eligibility (Yes/No)
+    2. Risk assessment
+    3. Recommended loan terms if eligible
+    4. Industry-specific insights
+    5. Decide if the plan is approved or not, depending on that create a Detailed 30-50 day actionable plan in Google Docs:
+       - If approved: Create a structured repayment strategy that optimizes cash flow
+       - If not approved: Outline specific steps to improve the application for approval
+       
+    For the action plan, include:
+    - Week-by-week breakdown of tasks
+    - Specific metrics to track
+    - Potential challenges and solutions
+    - Resources needed
+    My output should contain all the detailed analysis and also create the initial google doc using the tools available to you and provide the link to it.
+
+    """
+
+    response = client.models.generate_content(
+        model='gemini-2.0-pro-exp-02-05',
+        config=config,
+        contents=prompt
+    )
+
+    return response.text
+
+def main():
+    st.set_page_config(page_title="Business Loan Underwriting AI", page_icon="💼", layout="wide")
+
+    st.title("Business Loan Eligibility Checker")
+    st.subheader("Powered by Composio and Gemini")
+
+    client, config = initialize_ai()
+
+    with st.form("loan_application"):
+        col1, col2 = st.columns(2)
+
+        with col1:
+            business_name = st.text_input("Business Name")
+            industry = st.selectbox(
+                "Industry",
+                ["Retail", "Technology", "Manufacturing", "Healthcare", "Food & Beverage", 
+                 "Construction", "Professional Services", "Other"]
+            )
+            years_in_operation = st.number_input("Years in Operation", min_value=0.0, step=0.5)
+            annual_revenue = st.number_input("Annual Revenue ($)", min_value=0, step=1000)
+
+        with col2:
+            credit_score = st.slider("Business Credit Score", 300, 850, 650)
+            loan_amount = st.number_input("Desired Loan Amount ($)", min_value=0, step=1000)
+            current_debt = st.number_input("Current Outstanding Debt ($)", min_value=0, step=1000)
+
+        uploaded_files = st.file_uploader(
+            "Upload Supporting Documents (Optional)", 
+            accept_multiple_files=True,
+            help="Upload financial statements, tax returns, or business plans"
+        )
+
+        submit_button = st.form_submit_button("Evaluate Application")
+
+    if submit_button:
+        if not business_name or annual_revenue == 0 or loan_amount == 0:
+            st.error("Please fill in all required fields")
+            return
+
+        with st.spinner("Analyzing your application..."):
+            application_data = {
+                "business_name": business_name,
+                "industry": industry,
+                "years_in_operation": years_in_operation,
+                "annual_revenue": annual_revenue,
+                "credit_score": credit_score,
+                "loan_amount": loan_amount,
+                "current_debt": current_debt
+            }
+
+            evaluation = evaluate_loan_application(application_data, client, config)
+
+            st.success("Analysis Complete!")
+            st.markdown("### AI Evaluation")
+            st.write(evaluation)
+
+            # Add visual metrics
+            metrics = st.columns(3)
+            with metrics[0]:
+                debt_to_revenue = (current_debt / annual_revenue) * 100 if annual_revenue > 0 else 0
+                st.metric("Debt-to-Revenue Ratio", f"{debt_to_revenue:.1f}%")
+            with metrics[1]:
+                loan_to_revenue = (loan_amount / annual_revenue) * 100 if annual_revenue > 0 else 0
+                st.metric("Loan-to-Revenue Ratio", f"{loan_to_revenue:.1f}%")
+            with metrics[2]:
+                st.metric("Credit Score Rating", 
+                         "Excellent" if credit_score >= 750 else
+                         "Good" if credit_score >= 670 else
+                         "Fair" if credit_score >= 580 else
+                         "Poor")
+
+if __name__ == "__main__":
+    main()

python/examples/advanced_agents/loan_underwriter_agent/readme.md

@@ -0,0 +1,28 @@
+# Loan Underwriter Agent
+
+This guide provides detailed steps to run a Loan Underwriter Agent that leverages Composio and Gemini to create an Agent that underwrites Loans. Ensure you have Python 3.8 or higher installed.
+
+## Steps to Run
+
+**Navigate to the Project Directory:**
+Change to the directory where the `setup.sh`, `main.py`, `requirements.txt`, and `README.md` files are located. For example:
+```sh
+cd path/to/project/directory
+```
+
+### 1. Run the Setup File
+Make the setup.sh Script Executable (if necessary):
+On Linux or macOS, you might need to make the setup.sh script executable:
+```shell
+chmod +x setup.sh
+```
+Execute the setup.sh script to set up the environment and install dependencies:
+```shell
+./setup.sh
+```
+Now, fill in the `.env` file with your secrets.
+
+### 2. Run the Python Script
+```shell
+streamlit run cookbook/python-examples/advanced_agents/loan_underwriter_agent/main.py
+```

python/examples/advanced_agents/loan_underwriter_agent/requirements.txt

@@ -0,0 +1,3 @@
+composio-gemini
+streamlit
+python-dotenv

python/examples/advanced_agents/loan_underwriter_agent/setup.sh

@@ -0,0 +1,35 @@
+#!/bin/bash
+
+echo "Creating virtual environment..."
+python3 -m venv loan_agent
+
+# Activate the virtual environment
+echo "Activating virtual environment..."
+source loan_agent/bin/activate
+
+# Install libraries from requirements.txt 
+echo "Installing libraries from requirements.txt..."
+pip install -r requirements.txt
+
+# Copy env backup to .env file
+if [ -f ".env.example" ]; then
+    echo "Copying .env.example to .env..."
+    cp .env.example .env
+else
+    echo "No .env.example file found. Creating a new .env file..."
+    touch .env
+fi
+
+echo "Authenticating with Composio..."
+composio add googledocs
+# Prompt the user to enter the GEMINI_API_KEY
+read -p "Enter your GEMINI API KEY: " GEMINI_API_KEY
+
+# Update the .env file with the entered GEMINI_API_KEY
+sed -i "s/^GEMINI_API_KEY=.*$/GEMINI_API_KEY=$GEMINI_API_KEY/" .env
+
+echo "GEMINI_API_KEY has been set in the .env file"
+
+echo "Please fill in the .env file with any other necessary environment variables."
+
+echo "Setup completed successfully!"