mirror of
https://github.com/khoaliber/n8nworkflows.xyz.git
synced 2026-04-28 00:29:22 +00:00
creation
This commit is contained in:
nusquama
+284
@@ -0,0 +1,284 @@
|
||||
Client Feedback Collector & Analyzer (Form → AI Summary → Email + Social Draft)
|
||||
|
||||
https://n8nworkflows.xyz/workflows/client-feedback-collector---analyzer--form---ai-summary---email---social-draft--3910
|
||||
|
||||
|
||||
# Client Feedback Collector & Analyzer (Form → AI Summary → Email + Social Draft)
|
||||
|
||||
### 1. Workflow Overview
|
||||
|
||||
This workflow, named **Client Feedback Collector & Analyzer (Form → AI Summary → Email + Social Draft)**, automates the collection, analysis, and dissemination of client feedback. It is designed to intake raw feedback data submitted via a form (such as Tally, Typeform, or Google Forms), utilize AI to generate a structured summary and social media draft, and then distribute these outputs via email and Telegram (or other messaging integrations).
|
||||
|
||||
The workflow is organized into the following logical blocks:
|
||||
|
||||
- **1.1 Input Reception:** Captures client feedback data submitted through an HTTP webhook.
|
||||
- **1.2 AI Processing:** Prepares and sends the feedback to an AI service for analysis and content generation.
|
||||
- **1.3 Output Formatting:** Parses the AI response into distinct outputs for reporting and social media.
|
||||
- **1.4 Distribution:** Sends the summarized feedback report via email and the social media draft via Telegram.
|
||||
|
||||
This modular design supports easy adaptation, such as swapping messaging platforms or customizing AI prompts.
|
||||
|
||||
---
|
||||
|
||||
### 2. Block-by-Block Analysis
|
||||
|
||||
#### 2.1 Input Reception
|
||||
|
||||
- **Overview:**
|
||||
This block receives the client feedback data via an HTTP POST webhook endpoint. It expects a field named `feedback` carrying the client's textual input.
|
||||
|
||||
- **Nodes Involved:**
|
||||
- Receive Feedback
|
||||
|
||||
- **Node Details:**
|
||||
|
||||
- **Receive Feedback**
|
||||
- Type: Webhook node
|
||||
- Role: Entry point capturing POST requests from feedback forms.
|
||||
- Configuration:
|
||||
- HTTP Method: POST
|
||||
- Path: `/client-feedback`
|
||||
- No additional options enabled.
|
||||
- Inputs: None (start node)
|
||||
- Outputs: Passes JSON containing the client’s feedback field to the next node.
|
||||
- Edge cases:
|
||||
- Missing or empty `feedback` field results in default handling downstream.
|
||||
- Potential HTTP errors if the webhook URL is misconfigured or the form does not POST correctly.
|
||||
|
||||
#### 2.2 AI Processing
|
||||
|
||||
- **Overview:**
|
||||
This block constructs a prompt for the AI model based on the incoming feedback and sends it to the AI API to generate an analysis and social media draft.
|
||||
|
||||
- **Nodes Involved:**
|
||||
- Prepare AI Prompt
|
||||
- Analyze with AI
|
||||
|
||||
- **Node Details:**
|
||||
|
||||
- **Prepare AI Prompt**
|
||||
- Type: Function node
|
||||
- Role: Constructs a text prompt embedding the client's feedback for AI analysis.
|
||||
- Configuration:
|
||||
- JavaScript code extracts `feedback` from the input JSON; defaults to `"No feedback provided."` if absent.
|
||||
- Builds a prompt requesting:
|
||||
1. Summary of positive points
|
||||
2. Suggestions for improvement
|
||||
3. Short social media post based on positives
|
||||
- Key Expression:
|
||||
```js
|
||||
const feedback = $json.feedback || "No feedback provided.";
|
||||
return [{
|
||||
json: {
|
||||
prompt: `Analyze this client feedback: "${feedback}"\n\n1. Summarize the positive points.\n2. Suggest improvements.\n3. Generate a short social media post based on the positive elements.`
|
||||
}
|
||||
}];
|
||||
```
|
||||
- Inputs: JSON with `feedback` field from the webhook
|
||||
- Outputs: JSON with `prompt` string for the AI API
|
||||
- Edge cases: Empty or malformed feedback handled by default text.
|
||||
|
||||
- **Analyze with AI**
|
||||
- Type: HTTP Request node
|
||||
- Role: Sends the prepared prompt to the DeepSeek AI API (or compatible GPT service) and receives the generated text.
|
||||
- Configuration:
|
||||
- URL: `https://api.deepseek.com/generate`
|
||||
- HTTP Method: POST (implied)
|
||||
- Authentication: Predefined credential (DeepSeek API key or compatible)
|
||||
- Request body: JSON containing the prompt from the previous node
|
||||
- JSON parameters enabled for request formatting
|
||||
- Inputs: Prompt JSON from "Prepare AI Prompt"
|
||||
- Outputs: AI-generated response JSON
|
||||
- Version-specific: Uses n8n HTTP Request v2 node for better JSON handling
|
||||
- Edge cases:
|
||||
- API authentication failure
|
||||
- Network timeout
|
||||
- Unexpected API response structure
|
||||
|
||||
#### 2.3 Output Formatting
|
||||
|
||||
- **Overview:**
|
||||
This block processes the AI response to separate the analysis summary and the social media post draft.
|
||||
|
||||
- **Nodes Involved:**
|
||||
- Format AI Output
|
||||
|
||||
- **Node Details:**
|
||||
|
||||
- **Format AI Output**
|
||||
- Type: Function node
|
||||
- Role: Parses the AI response text to extract the report summary and social media post separately.
|
||||
- Configuration:
|
||||
- Extracts the AI output text from either `response` or `choices[0].text` fields.
|
||||
- Searches for the start of item "3." to split summary and post.
|
||||
- Returns two JSON fields: `report` (summary + improvements) and `post` (social media draft).
|
||||
- Key Expression:
|
||||
```js
|
||||
const output = $json.response || $json.choices?.[0]?.text || "No AI output.";
|
||||
const splitIndex = output.indexOf("3.");
|
||||
let summary = output;
|
||||
let post = "No post generated.";
|
||||
|
||||
if (splitIndex !== -1) {
|
||||
summary = output.substring(0, splitIndex).trim();
|
||||
post = output.substring(splitIndex).replace(/^3\./, "").trim();
|
||||
}
|
||||
|
||||
return [{
|
||||
json: {
|
||||
report: summary,
|
||||
post: post
|
||||
}
|
||||
}];
|
||||
```
|
||||
- Inputs: AI response JSON
|
||||
- Outputs: JSON with `report` and `post` fields for downstream nodes
|
||||
- Edge cases:
|
||||
- AI output missing or malformed results in default messages
|
||||
- Unexpected AI text formatting could break splitting logic
|
||||
|
||||
#### 2.4 Distribution
|
||||
|
||||
- **Overview:**
|
||||
This block sends the formatted feedback summary report via email and the social media draft via Telegram.
|
||||
|
||||
- **Nodes Involved:**
|
||||
- Send Feedback Report
|
||||
- Send Social Draft
|
||||
|
||||
- **Node Details:**
|
||||
|
||||
- **Send Feedback Report**
|
||||
- Type: Email Send node
|
||||
- Role: Emails the AI-generated feedback report to the specified team address.
|
||||
- Configuration:
|
||||
- Subject: "Client Feedback Summary"
|
||||
- To: `team@email.com` (placeholder; must be customized)
|
||||
- From: `your@email.com` (placeholder; must be customized)
|
||||
- Text body: Uses the `report` field from input JSON
|
||||
- SMTP credentials required (configured in n8n credentials)
|
||||
- Inputs: JSON with `report` text from "Format AI Output"
|
||||
- Outputs: None (terminal node)
|
||||
- Edge cases:
|
||||
- SMTP failures (auth, connection)
|
||||
- Invalid email addresses
|
||||
|
||||
- **Send Social Draft**
|
||||
- Type: Telegram node
|
||||
- Role: Sends the AI-generated social media post draft to a Telegram chat.
|
||||
- Configuration:
|
||||
- Text: Uses `post` field from input JSON
|
||||
- Chat ID: Placeholder `YOUR_TELEGRAM_CHAT_ID` (must be updated)
|
||||
- No additional fields set
|
||||
- Telegram credentials configured in n8n
|
||||
- Inputs: JSON with `post` text from "Format AI Output"
|
||||
- Outputs: None (terminal node)
|
||||
- Edge cases:
|
||||
- Incorrect chat ID or invalid Telegram credentials
|
||||
- Message length or formatting issues
|
||||
- Optional replacement with Slack, Buffer, or other integrations as per setup instructions
|
||||
|
||||
---
|
||||
|
||||
### 3. Summary Table
|
||||
|
||||
| Node Name | Node Type | Functional Role | Input Node(s) | Output Node(s) | Sticky Note |
|
||||
|----------------------|--------------------|-----------------------------------|---------------------|-----------------------------|----------------------------------------------------------------------------------------------------|
|
||||
| Receive Feedback | Webhook | Receives client feedback via POST | None | Prepare AI Prompt | Connect your form tool to this webhook using POST method; ensure it sends a `feedback` field. |
|
||||
| Prepare AI Prompt | Function | Builds AI prompt from feedback | Receive Feedback | Analyze with AI | Customize the prompt here to change tone or language if desired. |
|
||||
| Analyze with AI | HTTP Request | Sends prompt to AI API and gets response | Prepare AI Prompt | Format AI Output | Add your DeepSeek or compatible GPT API key in credentials for this node. |
|
||||
| Format AI Output | Function | Parses AI response into summary and post | Analyze with AI | Send Feedback Report, Send Social Draft | |
|
||||
| Send Feedback Report | Email Send | Emails the feedback summary | Format AI Output | None | Configure SMTP credentials and recipient email address here. |
|
||||
| Send Social Draft | Telegram | Sends social media draft via Telegram | Format AI Output | None | Replace Telegram with Slack, Buffer, or other service as needed; update chat ID and credentials. |
|
||||
|
||||
---
|
||||
|
||||
### 4. Reproducing the Workflow from Scratch
|
||||
|
||||
1. **Create Webhook Node: "Receive Feedback"**
|
||||
- Type: Webhook
|
||||
- HTTP Method: POST
|
||||
- Path: `client-feedback`
|
||||
- No additional options
|
||||
- This node will serve as the entry point for form submissions.
|
||||
|
||||
2. **Create Function Node: "Prepare AI Prompt"**
|
||||
- Connect output of "Receive Feedback" to this node.
|
||||
- Set code to:
|
||||
```js
|
||||
const feedback = $json.feedback || "No feedback provided.";
|
||||
return [{
|
||||
json: {
|
||||
prompt: `Analyze this client feedback: "${feedback}"\n\n1. Summarize the positive points.\n2. Suggest improvements.\n3. Generate a short social media post based on the positive elements.`
|
||||
}
|
||||
}];
|
||||
```
|
||||
- This node formats the prompt text for AI processing.
|
||||
|
||||
3. **Create HTTP Request Node: "Analyze with AI"**
|
||||
- Connect output of "Prepare AI Prompt" to this node.
|
||||
- URL: `https://api.deepseek.com/generate`
|
||||
- HTTP Method: POST (default)
|
||||
- Enable "JSON/RAW Parameters" for request body
|
||||
- Authentication: Select or create credential with your DeepSeek or compatible GPT API key
|
||||
- Payload: Use JSON with the `prompt` property passed from previous node
|
||||
- This node sends the prompt to the AI API and retrieves the result.
|
||||
|
||||
4. **Create Function Node: "Format AI Output"**
|
||||
- Connect output of "Analyze with AI" to this node.
|
||||
- Set code to:
|
||||
```js
|
||||
const output = $json.response || $json.choices?.[0]?.text || "No AI output.";
|
||||
const splitIndex = output.indexOf("3.");
|
||||
let summary = output;
|
||||
let post = "No post generated.";
|
||||
|
||||
if (splitIndex !== -1) {
|
||||
summary = output.substring(0, splitIndex).trim();
|
||||
post = output.substring(splitIndex).replace(/^3\./, "").trim();
|
||||
}
|
||||
|
||||
return [{
|
||||
json: {
|
||||
report: summary,
|
||||
post: post
|
||||
}
|
||||
}];
|
||||
```
|
||||
- This node splits the AI output into summary and social media draft.
|
||||
|
||||
5. **Create Email Send Node: "Send Feedback Report"**
|
||||
- Connect output of "Format AI Output" to this node.
|
||||
- Set "To Email" to your team's email address (e.g., `team@email.com`)
|
||||
- Set "From Email" to your sender address (e.g., `your@email.com`)
|
||||
- Subject: `Client Feedback Summary`
|
||||
- Text: Expression referencing `{{ $json["report"] }}`
|
||||
- Configure SMTP credentials in n8n for sending emails.
|
||||
|
||||
6. **Create Telegram Node: "Send Social Draft"**
|
||||
- Connect output of "Format AI Output" to this node.
|
||||
- Set "Chat ID" to your Telegram chat ID (`YOUR_TELEGRAM_CHAT_ID`)
|
||||
- Text: Expression referencing `{{ $json["post"] }}`
|
||||
- Configure Telegram credentials in n8n.
|
||||
- Optionally replace this node with Slack, Buffer, or other messaging nodes as preferred.
|
||||
|
||||
7. **Activate and Test**
|
||||
- Deploy the workflow inactive by default.
|
||||
- Send test POST requests with `feedback` field to the webhook URL.
|
||||
- Confirm email receipt and Telegram message delivery.
|
||||
|
||||
---
|
||||
|
||||
### 5. General Notes & Resources
|
||||
|
||||
| Note Content | Context or Link |
|
||||
|------------------------------------------------------------------------------------------------------------------|------------------------------------------------------------------------------------------------------------|
|
||||
| Estimated setup time is approximately 15 minutes for credential and webhook configuration. | Setup instructions in the workflow description. |
|
||||
| Sticky notes in the workflow provide visual guidance on key configuration steps such as API keys and webhook use.| Present within the n8n editor on respective nodes. |
|
||||
| Replace the Telegram node with Slack, Buffer, or other messaging services by changing node type and credentials. | Allows adaptation to different team communication platforms. |
|
||||
| DeepSeek API is used as an example GPT-compatible AI service; other providers can be configured similarly. | Requires corresponding API credentials and endpoint adjustments. |
|
||||
| Useful for creating actionable insights from customer feedback instantly, benefiting freelancers, educators, etc.| Enables marketing to react quickly via social drafts, and support teams to address improvements. |
|
||||
|
||||
---
|
||||
|
||||
This comprehensive analysis and reconstruction guide provides a clear, stepwise understanding of the workflow’s structure, rationale, node configurations, and integration points to facilitate seamless reproduction and customization.
|
||||
Reference in New Issue
Block a user