Async support, built-in polling helper & webhook callbacks for document processing

[shashank-518]

🐛 Problem Statement

I am integrating the PageIndex Python SDK into a production application and have successfully submitted documents using submit_document(). However, when polling for status via:

status = pi_client.get_document(doc_id)["status"]

The status remains "processing" for an extended period. The SDK provides no built-in mechanism to wait for completion — the only current option is a blocking while loop with time.sleep(), which is unsuitable for production-grade async backends.

---

💡 Proposed Improvements

1. ⚡ Async / Await Support

Provide an async-native client so status checks are non-blocking and integrate cleanly with FastAPI, asyncio, and other modern frameworks:

# Non-blocking, compatible with FastAPI / asyncio
status = await pi_client.get_document(doc_id)

---

2. 🔁 Built-in wait_until_completed() Helper

A high-level helper that handles polling internally, with configurable timeout and retry interval:

result = pi_client.wait_until_completed(
    doc_id,
    timeout=300,       # seconds
    poll_interval=5    # seconds
)

---

3. 🔔 Webhook / Callback Support

Supporting webhook notifications when processing completes would eliminate polling entirely — ideal for event-driven architectures:

pi_client.submit_document(
    file_path="report.pdf",
    webhook_url="https://myapp.com/callbacks/pageindex"
)
# PageIndex POSTs to the URL when processing completes
# Payload: { "doc_id": "...", "status": "completed" }

---

✅ Expected Benefit

These improvements would significantly enhance usability for production backend systems — especially those built on async frameworks. They would:

• Reduce boilerplate polling code
• Prevent resource waste from tight polling loops
• Align the SDK with modern Python ecosystem expectations

Is there already a recommended best practice for handling long-running document processing tasks with the current SDK? Happy to contribute a PR if the team is open to it.