Andrei's serverless function successfully generates a PDF (using pdf-lib) and logs indicate correct processing, but the PDF response (sent as base64 with appropriate headers) is not received by the client browser.

A suggested workaround involves generating the PDF externally (e.g., AWS Lambda with Puppeteer), obtaining its URL, using the HubSpot Files API to import the file from the URL into the HubSpot File Manager, and then potentially storing the resulting HubSpot File URL or ID on a CRM record.