Airtable DB vector search

Answers natural language questions about Airtable data using semantic search and AI. Accepts a user query, Airtable embedding collection name, embedding field, chat ID, and result limit. Returns a concise, relevant summary from Airtable. It generates OpenAI embeddings for the query, performs a vector search in Firestore, extracts unique Airtable record IDs, fetches those records, and uses OpenAI GPT to generate a context-aware answer.