Make.com Backend Data Pipeline Specialist

I am looking for an experienced Make.com specialist to build a 100% automated backend data pipeline for personal finance. The project involves integrating multiple APIs and automating data flows to ensure seamless data management. The ideal candidate will have a strong understanding of Make.com and its capabilities, as well as experience with API integration and data automation. Trigger (Fillout.com): A user completes a 3-page intake questionnaire on Fillout, processes a checkout payment via Stripe, and uploads a multi-page financial statement PDF (checking/savings transaction histories).Processing (OpenAI API): Make.com needs to capture this submission, grab the uploaded PDF file data, and pass it directly to the OpenAI API using the gpt-4o model (must be the native API connector module, not a generic assistant web-link). You will use a specific system prompt (provided upon hire) instructing the AI to analyze the raw transactions, categorize them, and extract structured data summary sheets.Delivery (Gmail): The text output generated by OpenAI must be mapped into a pre-formatted, clean HTML text email template and automatically sent back to the customer’s email address within 60 seconds of submission.Conditional Logic Gate (Data Routing): If a user selects a specific region from a dropdown menu AND inputs a financial metric value over a certain threshold, Make.com must simultaneously trigger an instant, formatted email alert summarizing the contact details to a separate partner email address.Strict Privacy & Security Requirements:You must configure the OpenAI API module to ensure that user data is explicitly passed dynamically and never cached, stored, or permitted to train public models.No internal databases or data stores (like Airtable or Make Data Stores) should retain the raw transaction history PDFs. Data must strictly flow through memory from the form to the email and immediately drop out of the cache.Clean error-handling must be implemented: If the API fails or a file cannot be parsed, an automated email must gracefully alert the admin.