"""Get article data from OpenAI with retry logic"""
    system_prompt = """Generate response in JSON format with these fields:
    1) 'snippet': A concise 3-5 sentence plain text summary (max 500-600 characters),
    2) 'content': Format the main article text in clean HTML preserving its structure,
    3) 'ts': Find and parse any publication timestamp (in ISO8601 format, e.g. 2024-01-21T00:00:00Z). Return empty string if no date found,
    4) 'author': Extract the article author's name if present, return empty string if not found.
    Content formatting requirements:
    - Include ONLY the main article content, remove navigation, footers, menus, links sections, etc
    - Use <h1>, <h2>, <h3> for article header hierarchy
    - Preserve paragraph structure with <p> tags
    - Keep and enhance formatting like <strong>, <em>, <code>, <blockquote>
    - Use <ul>/<li> for lists if present
    - Preserve any technical terms, code samples, or specific formatting
    - Make the article look professional and readable
    - Add proper spacing between elements
    - Be smart about identifying where the actual article content ends - exclude any supplementary sections, footers, related articles, etc"""

    data = {
        "model": "gpt-4o-mini",
        "messages": [
            {"role": "system", "content": system_prompt},
            {"role": "user", "content": content}
        ],
        "temperature": 0.3,
        "response_format": {"type": "json_object"}
    }