Automated Twitter Sentiment Analysis & Slack Alert Workflow

High-Level Summary

This workflow automatically monitors Twitter (X) for posts containing a specific keyword, analyzes their sentiment using OpenAI, and takes conditional actions based on the result. It is designed to help social media managers, marketers, and customer support teams track brand mentions or event-related tweets in real-time. The workflow fetches tweets via the Apify X Posts Search API, checks for duplicates in a Google Sheet, performs sentiment analysis with GPT-4o-mini, and then either generates a positive reply, sends a Slack alert for negative tweets, or logs neutral tweets for future reference. All data is stored in Google Sheets for auditability.

Step-by-Step Node Breakdown

1. Schedule Trigger — Runs the workflow every 6 hours. No auth required. The trigger is set to a 6-hour interval, ensuring regular monitoring without overwhelming API quotas.

2. Request for Twitter Post via Apify — Makes a POST request to the Apify API to search for tweets. API Key auth using query parameter token. It sends a JSON body with the query "LaraconIn", fetching 5 latest English tweets. This node will continue on error if the API fails.

3. Get Requested Post from Apify — Makes a GET request to retrieve the results from the Apify run. API Key auth using query parameter token. It fetches the dataset items from the most recent Apify run.

4. Set Field for Loop — Renames incoming JSON fields to standardized names (id, tweet_url, tweet_text). No auth. This ensures downstream nodes receive consistent field names regardless of API response structure.

5. Loop Over Items — Iterates over each tweet item. No auth. This split-in-batches node processes tweets one by one, enabling per-tweet actions.

6. Get Post Data — Queries Google Sheets to check if a tweet already exists. OAuth2 auth (Google Sheets). It looks up the tweet ID in column "ID" to detect duplicates.

7. If Duplicate — Checks if the tweet ID exists in the sheet. No auth. If the ID exists (true branch), the item is skipped to avoid duplicates. If not (false branch), the tweet proceeds to sentiment analysis.

8. OpenAI Chat Model — Provides the GPT-4o-mini language model for sentiment analysis. API Key auth (OpenAI). This node supplies the AI model to the Sentiment Analyst agent.

9. Structured Output Parser — Parses the sentiment result into structured JSON with a "category" field. No auth. It expects output in the format {"category": "neutral"} to ensure consistent data for the switch node.

10. Sentiment Analyst — Analyzes tweet text sentiment using GPT-4o-mini. No auth (uses the OpenAI Chat Model node). It classifies each tweet as Positive, Negative, or Neutral based on the system prompt.

11. Switch According Analyst — Routes tweets based on their sentiment category. No auth. Three outputs:

    • Positive: Routes to OpenAI for reply generation
    • Negative: Routes to Slack for alerting
    • Neutral: Routes directly to Google Sheets logging

12. OpenAI — Generates a user-style reply for positive tweets. API Key auth (OpenAI). It uses GPT-4o-mini to craft a short, friendly, emoji-free reply under 160 characters, directed at the original poster.

13. Send negative post message on slack — Sends a Slack message for negative tweets. OAuth2 auth (Slack). Posts to the #website-uptime channel with the tweet URL and a notification that action may be needed.

14. Add Post Data — Writes all tweet data back to Google Sheets. OAuth2 auth (Google Sheets). Appends or updates the row with Tweet ID, URL, text, sentiment, and the generated reply (if any). Matching is done on the ID column.

Setup Instructions

To use this workflow, you need accounts and credentials for the following services:

Prerequisites

1. Apify API Token (free tier available) — Sign up at apify.com and get your token from Settings > API & Integrations. Replace api_key in both HTTP Request URLs with your actual token.
2. OpenAI API Key — Sign up at platform.openai.com and create a project API key with access to GPT-4o-mini.
3. Google Sheets API — You need a Google Cloud Project with the Sheets API enabled. n8n provides OAuth2 credential setup for Google Sheets.
4. Slack App — Create a Slack app at api.slack.com with the chat:write scope and install it to your workspace. Channel ID C090F70N52M must be replaced with your actual channel.
5. n8n instance — Self-hosted or cloud (n8n.io).

Configuration

• Google Sheet: Create a sheet with columns: ID, TweetUrl, TweetText, Sentiment , Post Reply. Update the documentId and sheetName in both Google Sheets nodes.
• Search Query: In the "Request for Twitter Post via Apify" node, change query from "LaraconIn" to your desired keyword/hashtag.
• Schedule: Adjust the interval in the Schedule Trigger node as needed.

Use Cases & Variations

This workflow is highly adaptable. Consider these variations:

Use Cases

• Brand Monitoring: Track your company name or product mentions on Twitter, instantly alerting customer support to negative tweets.
• Event Tracking: Monitor hashtags for conferences (like #LaraconIn) to engage with attendees and gather feedback.
• Competitor Analysis: Track competitor product launches or campaigns to understand public sentiment.

Customizations

• Modify Sentiment Categories: Change the system prompt in Sentiment Analyst to detect urgency, spam, or specific topics.
• Add More Outputs: Route positive tweets to a Slack celebration channel, or negative tweets to a ticketing system like Jira.
• Different AI Models: Swap GPT-4o-mini for Claude or Gemini by changing the language model node.
• Expand Data Sources: Replace Apify with other Twitter scrapers or RSS feeds from social media listening tools.
• Add Human Review: Insert a manual approval step before sending replies to ensure brand safety.

{
  "name": "Twitter Monitoring Workflow",
  "nodes": [
    {
      "parameters": {
        "options": {}
      },
      "type": "n8n-nodes-base.splitInBatches",
      "typeVersion": 3,
      "position": [
        -60,
        210
      ],
      "id": "8fa04bf9-68e5-43b7-81f2-ee5f2355abd9",
      "name": "Loop Over Items"
    },
    {
      "parameters": {
        "modelId": {
          "__rl": true,
          "value": "gpt-4o-mini",
          "mode": "list",
          "cachedResultName": "GPT-4O-MINI"
        },
        "messages": {
          "values": [
            {
              "content": "={{ $('Set Field for Loop').item.json.tweet_text }}"
            },
            {
              "content": "=You are a professional Twitter social media user replying to the original poster (admin).\n\nAnalyze the sentiment of the tweet below. If the sentiment is positive, write a short, thoughtful reply **from a user's perspective**, directed to the person who posted it.\n\nThe reply should:\n- Show appreciation, agreement, or excitement\n- Be friendly, intelligent, and relevant to tech or AI fields\n- Use no emojis\n- Be under 160 characters\n- Address the original poster directly as a real person would\n\nTweet:\n\"\"\"\nuse a user prompt\n\"\"\"",
              "role": "system"
            }
          ]
        },
        "options": {}
      },
      "type": "@n8n/n8n-nodes-langchain.openAi",
      "typeVersion": 1.8,
      "position": [
        1380,
        -220
      ],
      "id": "dec03e00-a5fa-44eb-b269-b486259ac1aa",
      "name": "OpenAI",
      "credentials": {
        "openAiApi": {
          "id": "ga54eHkT3k6225Zq",
          "name": "OpenAi account"
        }
// ... truncated (copy to see full JSON)