How to Summarise Airtable with LlamaIndex

import os

import requests
from llama_index.core import VectorStoreIndex
from llama_index.core.schema import TextNode

BASE_URL = "https://api.enconvert.com/v2"
headers = {
    "X-API-Key": os.environ["ENCONVERT_API_KEY"],
    "Content-Type": "application/json",
}

article_urls = [url.strip() for url in os.environ["AIRTABLE_ARTICLE_URLS"].split(",")]  # public Airtable support or documentation page URLs

response = requests.post(
    f"{BASE_URL}/distill",
    headers=headers,
    json={
        "urls": article_urls,
        "schema": {
            "title": "The support article or documentation page title",
            "body": "The main article content, excluding navigation and related article links",
            "category": "The Airtable support category containing the article",
            "product_area": "The Airtable product, feature, or workflow covered",
            "last_updated": "The last updated date shown on the page, if available",
        },
    },
    timeout=120,
)

if response.status_code == 402:
    raise RuntimeError("Distill requires a paid plan")
response.raise_for_status()

# Each article is one unit — build TextNodes directly and skip the default splitter
nodes = []
for result in response.json()["results"]:
    if result["status"] != "completed":
        print(f"Skipped {result['url']}: {result['status']}")
        continue
    rq = result.get("render_quality", 1.0)
    if rq < 0.5:
        print(f"Low render quality for {result['url']}: {rq}")
        continue
    data = result.get("data") or {}
    nodes.append(
        TextNode(
            text=f"{data.get('title')}\n\n{data.get('body')}",
            id_=result["url"],
            metadata={
                "source_url": result["url"],
                "title": data.get("title"),
                "category": data.get("category"),
                "product_area": data.get("product_area"),
                "last_updated": data.get("last_updated"),
            },
        )
    )

if not nodes:
    raise RuntimeError("No Airtable pages passed the extraction checks")

# LlamaIndex's default embedding model and LLM read OPENAI_API_KEY from the environment
index = VectorStoreIndex(nodes)
query_engine = index.as_query_engine(similarity_top_k=3)
print(query_engine.query("Which Airtable product areas do these articles cover, and what does each one explain?"))