Automated Phone Interview Screening and Candidate Evaluation - n8n Workflow

• HR Managers & Recruiters: Seeking rapid, standardized screening processes for high-volume roles.


• Automation Specialists: Developers who need robust n8n workflow examples integrating advanced AI large language models (LLMs) and structured data parsing.


• Operations Teams: Users leveraging conversational AI platforms (like Vapi.ai) that generate detailed call transcripts.


• N8N Developers: Anyone looking for specialized n8n templates for data transformation and conditional logic using the LangChain n8n node.

Manually reviewing hundreds of phone screen transcripts is time-consuming and prone to human error. This n8n workflow solves this by creating a hands-off, objective evaluation system. By connecting your phone interview platform (via the webhook n8n trigger) to OpenAI's GPT-4o-mini model, the system autonomously applies a detailed checklist (e.g., license status, background, drug testing, availability) to every conversation. The result is a clean, structured JSON output detailing whether the candidate 'qualifies' and the specific reasoning, which is then immediately logged into Google Sheets. This highly specialized n8n automation eliminates manual review for initial screening, drastically speeding up the hiring pipeline and ensuring consistent application of job requirements. This complete n8n solution acts as a reliable filter for high-volume recruitment.

This n8n workflow executes a five-step automated sequence whenever a call transcript is finalized:

1. Trigger Reception: The n8n trigger, configured as a Webhook, waits for a POST request. This request is typically sent by a third-party conversational AI service (like Vapi.ai) immediately following an interview, containing the full call transcript and metadata.


2. Data Extraction: The Edit Fields n8n node extracts the nested transcript string from the raw webhook payload, making it easily accessible for the subsequent AI step.


3. AI Evaluation: The core of this n8n workflow lies in the Evaluate Candidate n8n node (a LangChain Agent). It passes the transcript and a highly detailed system message—which defines seven strict hiring criteria (for a driving job)—to the linked OpenAI Chat Model. The AI analyzes the transcript against these rules and attempts to output a structured JSON summary.


4. Output Formatting & Validation: The output from the initial evaluation is passed to a secondary LangChain Agent (Convert to JSON) and validated by the Structured Output Parser n8n node. This critical step ensures that the AI's response strictly adheres to the defined schema (Name, Phone, Qualifies [True/False], Reasoning), guaranteeing clean data quality.


5. Logging to Google Sheets: Finally, the Save to Google Sheets n8n node takes the validated, structured data fields and appends them as a new row to the designated Google Spreadsheet, completing the automated recruitment logging cycle. Every action in this n8n template is designed for efficiency and data integrity.

To deploy this powerful n8n workflow, follow these setup instructions:

1. Importing the N8N Template: Copy the provided JSON and import it into your n8n instance via the 'Workflows' > 'Import from JSON' option.


2. Configure Webhook Trigger: Click the 'Webhook' n8n node to view its unique URL. This URL must be configured as the endpoint in your phone interview platform (e.g., Vapi.ai) to send call completion data (like end-of-call-report).


3. Set Up OpenAI Credentials: Both the 'OpenAI Chat Model' and 'OpenAI Chat Model2' nodes require credentials. Create an OpenAI API key and configure the connection within the n8n interface. Ensure the gpt-4o-mini model is available for use.


4. Customize Evaluation Criteria (Optional but Recommended): Access the 'Evaluate Candidate' n8n node. Review and modify the System Message within the parameters to match your specific job requirements, disqualification rules, and desired output format.


5. Configure Google Sheets: Set up Google Sheets OAuth2 credentials in n8n. In the 'Save to Google Sheets' n8n node, select the correct Document ID and Sheet Name where you wish to log the candidate results.

This n8n workflow utilizes advanced LangChain agent nodes for intelligent processing:

Webhook (n8n trigger):
Function: Serves as the starting n8n trigger, waiting for a POST request containing the final transcript data from the interview system.
Key Configuration: HTTP Method is POST; listens on a unique path.
Edit Fields2 (Set n8n node):
Function: Extracts and cleans the deep-nested transcript path (body.message.artifact.transcript) into a simple field, streamlining subsequent data flow.
Evaluate Candidate (LangChain Agent n8n node):
Function: Performs the core AI logic. It uses the transcript as input and follows a strict, multi-point system message to determine candidate qualification (True/False).
Key Configuration: Uses a complex system prompt defining 7 specific criteria for a driver job. Connected to the gpt-4o-mini model.
Convert to JSON (LangChain Agent n8n node) & Structured Output Parser:
Function: These nodes work together to enforce a specific, clean JSON schema on the AI's output, ensuring consistency for database logging.
Key Configuration: The desired JSON schema (Name, Phone, Qualifies, Reasoning) is explicitly defined in the parser.
Save to Google Sheets (Google Sheets n8n node):
Function: Appends the structured evaluation results (extracted name, phone, cityState, qualifies, reasoning) as a new row to the specified spreadsheet.
Key Configuration: Operation set to Append. Data mapping uses expressions to pull fields from the validated JSON output.