Personalized AI Candidate Feedback and Learning Plans - n8n Workflow

HR Professionals and Recruiters looking to scale personalized communication.
Talent Acquisition teams focused on improving their employer brand and candidate experience.
n8n users seeking advanced n8n templates integrating PDF processing, Google Sheets, and large language models (LLMs).
Anyone needing an automated solution to provide constructive feedback to rejected applicants.

Maintaining a professional and empathetic candidate experience is crucial for modern HR teams. Generic rejection letters often lead to negative sentiment. This advanced n8n workflow solves this by providing highly customized, constructive feedback, even for rejected candidates. By pulling candidate data from Google Sheets, downloading and analyzing the resume PDF, and feeding all context into a GPT-4o model, the system automatically generates a personalized learning plan (in HTML format).

This single n8n template manages both positive (shortlisted/onboarding) and negative (rejected) outcomes, ensuring every candidate receives value. This automation demonstrates the power of the n8n platform to link disparate systems (Sheets, Drive, LLMs, Email) into one coherent, intelligent process. Every n8n node works together to deliver personalized communication at scale.

This complex n8n workflow begins with a manual trigger, allowing HR staff to execute the process when ready.

1. Data Ingestion (n8n trigger): The process starts with the 'When clicking ‘Execute workflow’' n8n trigger, which initiates the retrieval of structured candidate data (name, email, resume link, status, JD requirements) from a Google Sheets n8n node.


2. Resume Acquisition: The system uses the Google Drive n8n node (📥 Resume Downloader) to fetch the candidate's resume PDF based on the URL provided in the sheet.


3. Validation and Error Logging: An If n8n node (Resume File Check) validates if the resume was downloaded successfully. If not, the flow is redirected to the Error Logging Google Sheets n8n node.


4. Text Extraction: Valid files proceed to the PDF → Text Extractor n8n node, converting the resume PDF into raw text, making it readable for the AI.


5. Data Consolidation: The Candidate Data Builder Code n8n node merges the raw resume text with the original structured data from Google Sheets, forming a comprehensive input object for the LLM.


6. Flow Branching: The Shortlisted vs Rejected If n8n node splits the flow based on the candidate's 'Status' field (checking for 'Shortlisted').


7. AI Generation (Shortlisted): If 'Shortlisted', the workflow uses an Azure OpenAI n8n node (LLM Backend) and a Chain LLM n8n node to generate an encouraging HTML email detailing an 'Onboarding Plan,' including personalized learning recommendations.


8. AI Generation (Rejected): If 'Rejected', the workflow uses a separate Azure OpenAI n8n node (LLM Backend1) and a Chain LLM n8n node to generate a polite rejection email providing a constructive 'Learning Plan,' highlighting skill gaps and linking to relevant courses.


9. Communication: Finally, the respective Gmail n8n nodes (Candidate Mailer) send the personalized, HTML-formatted email to the candidate, completing the automated n8n workflow.

To deploy this comprehensive n8n workflow, follow these steps:

1. Import: Copy the provided JSON data and paste it into your n8n instance using the 'Import Workflow' function.


2. Google Sheets Setup: You need two Google Sheets credentials:

1. Google Drive Setup: Configure the 📥 Resume Downloader n8n node with a Google Drive OAuth2 credential that has read access to the location where the candidate resume links point.


2. AI Credentials: Set up the Azure OpenAI credential (azureOpenAiApi) used by the LLM Backend n8n node. Ensure the model (gpt-4o-mini) is available via your service.


3. Gmail Configuration: Configure the Candidate Mailer – Shortlisted and Candidate Mailer – Rejected Gmail n8n nodes with your sending Gmail or Google Workspace OAuth2 credentials. The 'sendTo' parameter dynamically pulls the email address from the initial Sheets data.


4. Test: Execute the n8n trigger manually to process a test candidate row and verify that the personalized emails are generated and sent correctly.

When clicking ‘Execute workflow’ (n8n trigger): Serves as the starting point for running the n8n workflow on demand.
📑 Candidate Data Fetch (Google Sheets n8n node): Retrieves structured data (status, resume link, job requirements) for the candidate being processed.
📥 Resume Downloader (Google Drive n8n node): Downloads the PDF resume linked in the Google Sheet data.
Resume File Check (If n8n node): Adds resilience to the n8n workflow by checking for a valid file download before proceeding to text extraction.
PDF → Text Extractor (Extract From File n8n node): Converts the downloaded resume PDF content into clean, readable text, which is vital context for the AI.
Candidate Data Builder (Code n8n node): A crucial step that combines the structured Sheets data and the unstructured resume text into one enriched JSON object for the subsequent AI processing steps.
Shortlisted vs Rejected (If n8n node): Manages the core branching logic, directing the item down the path appropriate for the candidate's application status.
LLM Backend / LLM Backend1 (Azure OpenAI Chat n8n node): Provides the necessary API connection to GPT-4o-mini, acting as the foundation for the generative AI tasks within the n8n template.
Congrats + Onboarding Plan / Polite Rejection + Learning Plan (Chain LLM n8n node): These use precise system prompts to analyze the unified candidate data and generate personalized, HTML-formatted communication (congratulations or constructive rejection) with skill gap analysis and real course recommendations.
Candidate Mailer (Gmail n8n node): Sends the final, personalized HTML email communication to the candidate, utilizing the dynamic email address fetched at the start of the n8n workflow.