Loss Run Report Processing System

Overview

This project is a Django-based application designed to process loss run reports from PDF files. It extracts relevant data, validates it, and stores it in a database. The system also uses Celery for background task processing and integrates with the OpenAI API for claim summarization and sentiment analysis.

Features

• PDF Data Extraction: Extracts data from loss run reports in PDF format.
• Data Validation: Validates extracted data using Django REST framework serializers.
• Background Processing: Uses Celery for asynchronous task processing.
• OpenAI Integration: Utilizes OpenAI's GPT models for claim summarization and sentiment analysis.
• Multi-tenancy: Support for multiple tenants, allowing businesses to manage their own wallets and customers.
• Shared Database & Shared Tables: All tenants share the same tables, simplifying database management.
• Tenant-Aware Queries: Middleware dynamically identifies the tenant and filters queries by tenant_id to ensure data isolation.
• Single Database, Single Schema: No need for separate schemas for each tenant.
• Scalability: This architecture scales well for tenants that don't require significant isolation.
• Simplified Migrations: Apply migrations once, and they will be applied to all tenants sharing the same tables.

Technical Details

Stack

• Backend: Django, Django REST framework
• Database: PostgreSQL
• Task Queue: Celery with Redis as the broker
• Machine Learning: OpenAI API for NLP tasks
• Containerization: Docker, Docker Compose
• Reverse Proxy: Nginx

Directory Structure

loss_run_project/
│
├── loss_run/
│   ├── __init__.py
│   ├── models.py
│   ├── serializers.py
│   ├── tasks.py
│   ├── utils.py
│   ├── views.py
│   └── handlers
│
├── claim_flow/
│   ├── __init__.py
│   ├── settings.py
│   ├── urls.py
│   └── wsgi.py
│
├── docker-compose.yml
├── Dockerfile
├── nginx.conf
├── requirements.txt
└── README.md

Quickstart

Prerequisites

• Docker
• Docker Compose
• Python 3.10
• pip

Installation

Clone the Repository:

git clone https://github.com/akanuragkumar/ClaimFlowAI.git
cd ClaimFlowAI

Install Dependencies:

pip install -r requirements.txt

Build and Run Containers:

docker-compose up --build

Access the Application:

Open your browser and navigate to http://localhost.

Stop the Containers:

docker-compose down

Environment Variables

Ensure you have the following environment variables set:

• DJANGO_SETTINGS_MODULE: Path to your Django settings module.
• DATABASE_URL: URL for your PostgreSQL database.
• REDIS_URL: URL for your Redis instance.
• OPENAI_API_KEY: Your OpenAI API key.

Database Migrations

Run the following command to apply database migrations:

docker-compose run web python manage.py migrate

Implementation Details

Data Extraction

The utils.py module contains functions to extract data from PDF files using pdfplumber and regular expressions.

Data Validation

The serializers.py module contains serializers to validate the extracted data before saving it to the database.

Background Tasks

The tasks.py module contains Celery tasks for data enrichment, sentiment analysis, and report generation.