Description
Summary I’m looking for an experienced n8n developer with strong OpenAI/LLM and REST API integration experience to build an end-to-end AI-assisted curriculum workflow for an interactive Bible-learning application. The backend API is already complete and documented in Postman. I will also provide access to the n8n project, Google Sheets used for staging/review, sample data structures, and curriculum-generation requirements. You may start the n8n implementation from scratch. The workflow should: 1. Read course input from Google Sheets. 2. Use AI to generate a course blueprint. 3. Break the blueprint into separate course, section, unit, and lesson records. 4. Save generated records into Google Sheets for human review. 5. Allow records to be edited and marked Approved before upload. 6. Generate full lesson content for approved lesson outlines. 7. Save each generated lesson immediately so completed work is not lost if a later lesson fails. 8. Find approved records and upload them through the existing REST API. 9. Upload records in the correct hierarchy: Course → Section → Unit → Lesson. 10. Save API-generated IDs back to the staging sheets so child records can reference their correct parents. 11. Track generation status, review status, upload status, errors, and timestamps. 12. Support safe retries so a failed lesson or API request does not require restarting an entire course. The spreadsheet will act as the initial staging and review environment. I want to be able to review generated content, edit titles or descriptions manually, approve or reject individual records, regenerate an individual lesson if needed, trigger upload only for approved content, and see whether each record has been uploaded successfully. Review status and API/upload status should be tracked separately. The REST API is already built. I will provide the Postman collection, development API credentials, endpoints for creating courses, sections, units, lessons, etc., and sample request/response payloads. The developer should integrate with the existing API rather than redesign the application database. Reliability is important. The workflow should be designed so that if lesson 27 fails, lessons 1–26 remain saved. Only failed or incomplete lessons should need to be retried. Successfully uploaded records should not be duplicated when a workflow is rerun. Parent IDs should be available before child records are uploaded. LLM/API errors should be logged clearly, and long-running generation should be able to resume safely. Required skills: n8n OpenAI API / LLM integration REST API integration JavaScript JSON and data transformation Google Sheets Workflow automation Error handling and retries HTTP authentication Looping/batching Sub-workflows Experience with Firebase/Firestore, Postman, structured AI output, or human-in-the-loop workflows is a plus. I will provide the curriculum subject matter, course titles, structure requirements, theological/content guidance, prompt requirements, API/Postman collection, n8n workspace access, Google Sheets, and sample records. You do not need to be a theology or curriculum expert. The initial deliverable should be a working end-to-end system that can take one test course through: Course input → AI blueprint → staged records → human approval → lesson generation → approval → API upload → returned application IDs saved back to staging The workflow should be well organized, maintainable, and documented. Please begin your proposal with “Curriculum Workflow” so I know you reviewed the description.