Airtable Dynamic AI Record Updater: Webhook-Triggered PDF Processing → Dynamic Field Population via LLM

Problem:

A business storing PDF-based data in Airtable had no automated way to process incoming files, extract relevant field values, and update the correct Airtable records dynamically, every new document required manual reading, data extraction, and field-by-field record updates, making large-scale data management unsustainable.

Solution:

Built a complex n8n Airtable automation system triggered by Airtable webhooks, the system retrieves the table schema and prompt fields, captures the webhook payload and event type, filters for valid rows with input, and runs two parallel processing paths via Items Loop: Path 1 fetches the PDF file, extracts text content, passes it through a Dynamically Generated Prompt using OpenAI to extract field values, and updates the Airtable record directly; Path 2 handles value decomposition using LLM, fetching references, extracting from file, generating field updates via OpenAI, and updating records using an Add New ID to Payload pattern, with a FieldsChanged webhook trigger ensuring only modified records trigger reprocessing.

Result:

Delivered a fully automated Airtable record enrichment system that detects document changes, processes PDF content through AI, and populates dynamic field values across records automatically, eliminating all manual data extraction and entry from document-heavy Airtable workflows.