AVO API - AI-Powered Guideline Processing

Minimal backend API for automated guideline processing with AI-powered summarization and checklist generation.

System Architecture

graph TD;
    A[Client] -->|POST /jobs| B[Django API];
    B -->|200ms response| A;
    B -->|Queue Task| C[Redis Queue];
    C -->|FIFO| D[Celery Worker];
    D -->|Step 1| E[GPT Summarize];
    E -->|Step 2| F[GPT Checklist];
    F -->|Save Results| G[PostgreSQL];
    A -->|GET /jobs/id| B;
    B -->|Query Status| G;
    G -->|Return Result| B;
    B -->|Status and Result| A;
    
    classDef client fill:#e1f5fe,stroke:#01579b,stroke-width:2px;
    classDef api fill:#f3e5f5,stroke:#4a148c,stroke-width:2px;
    classDef queue fill:#fff3e0,stroke:#e65100,stroke-width:2px;
    classDef worker fill:#e8f5e8,stroke:#1b5e20,stroke-width:2px;
    classDef gpt fill:#fce4ec,stroke:#880e4f,stroke-width:2px;
    classDef db fill:#e3f2fd,stroke:#0d47a1,stroke-width:2px;
    
    class A client;
    class B api;
    class C queue;
    class D worker;
    class E,F gpt;
    class G db;

Loading

Quick Start

# Clone and start services
git clone https://github.com/yhskgo/avo-api.git
cd avo-api
cp .env.example .env
# Edit .env with your OPENAI_API_KEY
docker compose up --build

# Run tests (6 tests, comprehensive coverage)
docker compose --profile test run test

# API Documentation
http://localhost:8000/api/docs/

API Endpoints

• POST /api/jobs → Returns event_id in <200ms
• GET /api/jobs/{event_id} → Job status and results
• GET /api/schema/ → OpenAPI specification
• GET /api/docs/ → Interactive API documentation

Design Choices

Tech Stack: Django + Celery + Redis + PostgreSQL for:

• Sub-200ms response: Immediate job queuing with async processing
• FIFO guarantee: Redis queues ensure order preservation
• Scalability: Horizontal worker scaling with Celery
• Reliability: PostgreSQL for persistent job state

Two-Stage GPT Chain:

1. Summarize guidelines using GPT-4o-mini/GPT-3.5-turbo
2. Generate actionable checklist from summary

AI Tools Used:

• IDE:VSCode:Python extensions for development environment and code debugging
• Claude: Complete system architecture, code structure, testing framework, and Docker deployment setup
• Mermaid: Interactive system architecture diagram
• OpenAPI: Automatic API documentation with drf-spectacular
• GPT Integration: Content processing pipeline with robust fallback handling

Built with extensive AI assistance for rapid, production-ready development.