AI Conversational Interview Agent using Custom Forms and Redis Session Memory - n8n Workflow

• User Researchers and CX Professionals needing scalable interview solutions.


• Developers interested in creating interactive, stateful AI chatbots.


• Teams utilizing n8n for complex AI automation flows.


• Users looking for advanced examples of the n8n form trigger and loop control.

Conducting scalable user research interviews is often cost-prohibitive and time-intensive. This powerful n8n workflow solves this by deploying a sophisticated AI Interview Agent that dynamically asks follow-up questions based on user input, creating a never-ending interview loop until the user explicitly asks to stop. The entire conversation transcript is recorded using a Redis session manager to maintain context, ensuring a fluid interview experience. This comprehensive n8n template demonstrates how to combine the conversational power of AI agents with the easy data capture of the n8n form trigger, storing the results in Google Sheets for analysis. Every step, from the initial form submission to the final transcript display, is managed by an n8n node, showcasing the platform’s potential for complex, user-facing applications.

This stateful n8n workflow operates in a continuous loop powered by n8n Forms and an AI Agent:

1. Start Trigger: The process begins with the Start Interview n8n trigger (Form Trigger node). The user submits their name, initiating the session.


2. Session Initialization: A unique UUID is generated, and the Create Session Redis n8n node initializes a new session key for storing the transcript. The initial user input is saved.


3. AI Question Generation: The core loop starts at the AI Researcher (Langchain Agent) n8n node. This agent, configured with system prompts and backed by a Groq Chat Model, analyzes the conversation history (maintained via Window Buffer Memory) and generates the next question, outputting a structured JSON response (including a stop_interview boolean).


4. Loop Control: The Parse Response n8n node processes the AI's output. The Stop Interview? If n8n node checks the boolean flag. If False, the loop continues.


5. User Input: The Get Answer n8n node presents the AI's generated question as a new form to the user, collecting the response.


6. Session Update: The user's answer is formatted (Generate Row) and pushed to the persistent Redis session using the Update Session1 n8n node, ensuring the AI agent retains context for the next turn.


7. Loop Continuation: The user's answer is passed back to the AI Researcher n8n node (Send Reply To Agent), restarting the conversation loop.


8. Workflow Termination: If the AI detects the user wishes to stop, the Stop Interview? n8n node routes the flow to the cleanup path. The session is closed, the memory is deleted, and the user is redirected via a custom redirect URL in the Redirect to Completion Screen n8n node.


9. Data Persistence & Transcript: The full session is fetched from Redis and uploaded row-by-row to a Google Sheet using the Save to Google Sheet n8n node for later analysis. A separate Webhook n8n trigger captures the redirected request and dynamically displays the session transcript to the user as a completion screen.

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

1. Import: Copy the provided JSON data and import it directly into your n8n instance as a new workflow.


2. Credentials: You must configure the following application credentials:

1. Webhook Configuration: For the transcript display functionality, you must copy the URL generated by the second Webhook n8n trigger (the one at the far end of the workflow, path /ai-interview-transcripts/:session_id). Paste this full URL into the 'Redirect to Completion Screen' Form n8n node's redirectUrl parameter, replacing the placeholder values ( and ).


2. Activation: Once credentials and redirects are set, activate the Start Interview n8n trigger and access its form URL to begin testing the conversational AI.

Start Interview (Form Trigger): This n8n trigger starts the entire n8n workflow when a user fills out the initial form (asking for their name). Key configuration includes the form title and a required name field.
UUID (Crypto): Generates a unique identifier used to manage the Redis session key, ensuring isolation between multiple concurrent interviews.
Create Session (Redis n8n node): Initializes the Redis key (set to expire after 24 hours) where the entire conversation transcript will be stored in list format.
AI Researcher (Langchain Agent n8n node): The brain of the automation. It takes the user's answer and, utilizing the Groq Chat Model and Window Buffer Memory, generates the next logical interview question based on the detailed system prompt. This ensures the conversational flow is maintained.
Groq Chat Model (Langchain LLM): Provides the large language model backend (llama-3.2) necessary for the advanced reasoning and question generation required by the AI agent.
Stop Interview? (If n8n node): Essential flow control logic. It parses the AI's JSON output to determine if the continuous n8n workflow loop should be broken (e.g., if the user typed 'stop').
Get Answer (Form n8n node): A crucial element of the loop. It presents the AI's newly generated question to the user as a new form field, capturing the reply and advancing the conversational process.
Update Session (Redis n8n node): Appends the latest question/answer pair to the Redis list, persistently updating the memory of the conversation throughout the n8n workflow.
Redirect to Completion Screen (Form n8n node): On interview termination, this node redirects the user's browser to the separate transcript webhook URL.
Save to Google Sheet (Google Sheets n8n node): Used for permanent storage. It receives the entire session data and appends structured rows to a designated sheet for data analysis. This ensures the results of the n8n workflow are captured effectively.