RiskScope AI - Comprehensive Underwriting Agent

RiskScope AI is an advanced risk assessment and underwriting tool designed to automate the due diligence process for insurance and financial analysis. By leveraging the Firecrawl API for autonomous web research, the application aggregates critical data points across firmographics, financial health, operational stability, and adverse signals to generate a comprehensive risk score for target companies.

Features

• Automated Data Extraction: Utilizes AI agents to research companies in real-time, gathering data from diverse public sources.
• Comprehensive Risk Scoring: Implements a sophisticated weighted scoring algorithm to quantify risk:
  • Firmographics (15%)
  • Financial Health (30%)
  • Operational Risk (25%)
  • Adverse Signals (30%)
• Interactive Dashboard: A user-friendly Streamlit interface for inputting targets and visualizing risk metrics.
• Detailed Analytics: Provides granular breakdowns of risk factors, including leadership stability, credit ratings, revenue trends, and regulatory compliance.
• Cost-Efficient Architecture: Features an optimized single-call extraction mode to maximize data retrieval while minimizing API credit usage.

Technology Stack

• Frontend: Streamlit
• Language: Python 3.x
• Data Extraction: Firecrawl API
• Visualization: Plotly
• Logging: Loguru

Prerequisites

Before running the application, ensure you have the following installed:

• Python 3.8 or higher
• pip (Python package installer)

You will also need a valid API key from Firecrawl.

Installation

1. Clone the repository Navigate to the project directory on your local machine.

2. Create a virtual environment (Recommended) It is best practice to run Python applications in an isolated environment.

   python -m venv venv
   source venv/bin/activate  # On Windows use: venv\Scripts\activate

3. Install dependencies Install the required Python packages listed in requirements.txt.

   pip install -r requirements.txt

Configuration

The application requires environment variables to function correctly.

1. Locate the .env.template file in the root directory.
2. Create a copy of this file and name it .env.
3. Open the .env file and enter your Firecrawl API key.

FIRECRAWL_API_KEY=your_api_key_here

Usage

To start the application, use the Streamlit CLI command:

streamlit run app.py

Once started, the application will open in your default web browser (typically at http://localhost:8501). Enter the name of the company you wish to analyze to begin the risk assessment process.

Risk Scoring Methodology

The application evaluates companies based on four primary risk pillars:

1. Firmographics (15%): assesses business maturity, entity type, and structural stability.
2. Financial Health (30%): evaluates solvency/liquidity ratios, revenue trends, and credit information.
3. Operational Risk (25%): analyzes leadership stability (C-suite changes), certifications, and employee sentiment.
4. Adverse Signals (30%): scans for regulatory actions, legal disputes, and negative media coverage.

Project Structure

• app.py: Main application entry point containing the Streamlit UI, risk scoring logic, and data processing.
• requirements.txt: List of Python dependencies.
• logs/: Directory where application logs are stored.
• .env: Configuration file for sensitive credentials (not committed to version control).