Workflow Templates

🧩 Workflow: Process Tour PDF from Google Drive to Pinecone Vector DB with OpenAI Embeddings Overview This workflow automates the process of extracting tour information from PDF files stored in a Google Drive folder, processes and vectorizes the extracted data, and stores it in a Pinecone vector database for efficient querying. This is especially useful for building AI-powered search or recommendation systems for travel packages. Setup: Prerequisites A folder in Google Drive with PDF tour package brochures. Pinecone account + API key OpenAI API key n8n cloud or self-hosted instance Workflow Setup Steps Trigger Manual Trigger (When clicking 'Test workflow'): Used for manual testing and execution of the workflow. Google Drive Integration Step 1: Store Tour Packages in PDF Format Upload your curated tour packages containing the tours, activities and sight-seeings in PDF format into a designated Google Drive folder. Step 2: Search Folder Node: PDF Tour Package Folder (Google Drive) This node searches the designated folder for files (filter by MIME type = application/pdf if needed). Step 3: Download PDFs Node: Download Package Files (Google Drive) Downloads each matching PDF file found in the previous step. Process Each PDF File Step 4: Loop Through Files Node: Loop Over each PDF file Iterates through each downloaded PDF file to extract, clean, split, and embed. Data Preparation & Embedding Step 5: Data Loader Node: Data Loader Reads each PDF’s content using a compatible loader. It passes clean raw text to the next node. Often integrated with document loaders like pdf-loader, Unstructured, or pdfplumber. Step 6: Recursive Text Splitter Node: Recursive Character Text Splitter Splits large chunks of text into manageable segments using overlapping window logic (e.g., 500 tokens with 50 token overlap). This ensures contextual preservation for long documents during embedding. Step 7: Generate Embeddings Node: Embeddings OpenAI Uses text-embedding-3-small model to vectorize the split chunks. Outputs vector representations for each content chunk. Store in Pinecone Step 8: Pinecone Vector Store Node: Pinecone Vector Store - Store... Stores each embedding along with its metadata (source PDF name, chunk ID, etc.). This becomes the basis for fast, semantic search via RAG workflows or agents. 🛠️ Tools & Nodes Used Google Drive (Search & Download) Searches for all PDF files in a specified Google Drive folder. Downloads each file for processing. SplitInBatches (Loop Over Items) Loops through each file found in the folder, ensuring each is processed individually. Default Data Loader (LangChain) Reads and extracts text from the PDF files. Recursive Character Text Splitter (LangChain) Splits the extracted text into manageable chunks for embedding. OpenAI Embeddings (LangChain) Converts each text chunk into a vector using OpenAI’s embedding model. Pinecone Vector Store (LangChain) Stores the resulting vectors in a Pinecone index for fast similarity search and querying. 🔗 Workflow Steps Explained Trigger: The workflow starts manually for testing or can be scheduled. Google Drive Search: Finds all PDF files in the specified folder. Loop Over Files: Each file is processed one at a time using the SplitInBatches node. Download File: Downloads the current PDF file from Google Drive. Extract Text: The Default Data Loader node reads the PDF and extracts its text content. *Text Splitting: * The Recursive Character Text Splitter breaks the text into chunks (e.g., 1000 characters with 50 overlap) to optimize embedding quality. **Vectorization: **Each chunk is sent to the OpenAI Embeddings node to generate vector representations. Store in Pinecone: The vectors are inserted into a Pinecone index, making them available for semantic search and recommendations. 🚀 What Can Be Improved in the Next Version? *Error Handling: * Add error handling nodes to manage failed downloads or extraction issues gracefully. File Type Filtering: Ensure only PDF files are processed by adding a filter node. Metadata Storage: Store additional metadata (e.g., file name, tour ID) alongside vectors in Pinecone for richer search results. *Parallel Processing: * Optimize for large folders by processing multiple files in parallel (with care for API rate limits). Automated Triggers: Replace manual trigger with a time-based or webhook trigger for full automation. Data Validation: Add checks to ensure extracted text contains valid tour data before vectorization. User Feedback: Integrate notifications (e.g., email or Slack) to inform when processing is complete or if issues arise. 💡 Summary This workflow demonstrates how n8n can orchestrate a powerful AI data pipeline using Google Drive, LangChain, OpenAI, and Pinecone. It’s a great foundation for building intelligent search or recommendation features for travel and tour data. Feel free to ask for more details or share your improvements! Let me know if you want to see a specific part of the workflow or need help with a particular node!

🤖🥗 Telegram Nutrition AI Assistant (Alternative to Cal AI App) > AI-powered nutrition assistant for Telegram — log meals, set goals, and get personalized daily reports with Google Sheets integration. 📋 Description This n8n template creates a Telegram-based Nutrition AI Assistant 🥑🔥 designed as an open-source alternative to the Cal AI mobile app. It allows users to interact with an AI agent via text, voice, or images to track meals, calculate macros, and monitor nutrition goals directly from Telegram. The system integrates Google Sheets as the database, handling both user profiles and meal logs, while leveraging Gemini AI for natural conversation, food recognition, and daily progress reports. ✨ Key Features 💬 Multi-input support: Text, voice messages (transcribed), and food images (AI analysis). 📊 Macro calculation: Automatic estimation of calories, proteins, carbs, and fats. 📝 User-friendly registration: Simple onboarding without storing personal health data (no weight/height required). 🎯 Goal tracking: Users can set and update calorie and protein targets. 📈 Daily reports: Personalized progress messages with visual progress bars. 🗂 Google Sheets integration: Profile table for user targets. Meals table for food logs. 🔄 Advanced n8n nodes: Includes use of Merge, Subworkflow, and Code nodes for data processing and report generation. 💡 Acknowledgment Inspired by the Cal AI concept 💡 — this template demonstrates how to reproduce its main functionality with n8n, Telegram, and AI agents as a flexible, open-source automation workflow. 🏷 Tags telegram ai-assistant nutrition meal-tracking google-sheets food-logging voice-transcription image-analysis daily-reports n8n-template merge-node subworkflow-node code-node telegram-trigger google-gemini 💼 Use Case Use this template if you want to: 🥗 Log meals using text, images, or voice messages. 📊 Track nutrition goals (calories, proteins) with daily progress updates. 🤖 Provide a chat-based nutrition assistant without building a full app. 🗂 Store structured nutrition data in Google Sheets for easy access and analysis. 💬 Example User Interactions 📸 User sends a photo of a meal → AI analyzes the food and logs calories/macros. 🎤 User sends a voice message → AI transcribes and logs the meal. ⌨️ User types “report” → AI returns a daily nutrition summary with progress bars. 🥅 User says “update my protein goal” → AI updates profile in Google Sheets. 🔑 Required Credentials Telegram Bot API (Bot Token) Google Sheets API credentials AI Provider API (Google Gemini or compatible LLM) ⚙️ Setup Instructions 🗂 Create two Google Sheets tables: Profile: User_ID, Name, Calories_target, Protein_target Meals: User_ID, Date, Meal_description, Calories, Proteins, Carbs, Fats 🔌 Configure the Telegram Trigger with your bot token. 🤖 Connect your AI provider credentials (Gemini recommended). 📑 Connect Google Sheets with your credentials. ▶️ Deploy the workflow in n8n. 🎯 Start interacting with your nutrition assistant via Telegram. 📌 Extra Notes 🟩 Green section: Handles Telegram trigger and user check. 🟥 Red section: Registers new users and sets goals. 🟦 Blue section: Processes text, voice, and images. 🟨 Yellow section: Generates nutrition reports. 🟪 Purple section: Main AI agent controlling tools and logic. 💡 Need Assistance? If you’d like help customizing or extending this workflow, feel free to reach out: 📧 Email: johnsilva11031@gmail.com 🔗 LinkedIn: John Alejandro Silva Rodríguez

This n8n workflow demonstrates how to build a simple uptime monitoring service using scheduled triggers. Useful for webmasters with a handful of sites who want a cost-effective solution without the need for all the bells and whistles. How it works Scheduled trigger reads a list of website urls in a Google Sheet every 5 minutes Each website url is checked using the HTTP node which determines if the website is either in the UP or DOWN state. An email and Slack message are sent for websites which are in the DOWN state. The Google Sheet is updated with the website's state and a log created. Logs can be used to determine total % of UP and DOWN time over a period. Requirements Google Sheet for storing websites to monitor and their states Gmail for email alerts Slack for channel alerts Customising the workflow Don't use Google Sheets? This can easily be exchanged with Excel or Airtable.

Who is this for? This workflow is built for newsletter writers, marketers, content creators, or anyone who curates and summarizes web articles. It’s especially helpful for virtual assistants and founders who need to quickly turn web content into digestible, branded newsletters using AI. What problem is this workflow solving? Manually reading, summarizing, and formatting multiple articles into a newsletter takes time and focus. This workflow automates the process using Dumpling AI for crawling, GPT-4o for summarization, and Gmail for delivery—so you can go from raw URLs to a polished email in minutes. What this workflow does Starts manually (can also be scheduled) Reads a list of article URLs from Google Sheets Sends URLs to Dumpling AI to crawl and extract content Splits each article into a single item for processing Uses a Code node to clean and structure article data Uses an Edit Fields node to merge articles into one JSON block GPT-4o summarizes and generates HTML content for the newsletter Sends the formatted newsletter via Gmail Setup Google Sheets Create a sheet with a column (A) for article URLs Update the Read URLs from Google Sheet node to use your Sheet ID and tab name Connect your Google account in the credentials Dumpling AI Sign up at https://app.dumplingai.com Create an agent for web crawling under /crawl Add your Dumpling API key in the HTTP headers of the Crawl Content with Dumpling AI node Split Node Breaks apart the array of articles from Dumpling AI so each article is processed individually Code Node Structures each article as JSON with title, url, and cleaned text content Edit Fields Node Gathers all structured articles back into a single JSON array to prepare for AI summarization OpenAI (GPT-4o) Processes the article list and returns a formatted subject line and HTML newsletter content Gmail Connect your Gmail account to send the AI-generated newsletter to your inbox or team Update the recipient field in the Send HTML Email via Gmail node How to customize this workflow to your needs Replace the manual trigger with a Schedule node to send newsletters weekly Modify the GPT-4o prompt to change tone (e.g., more professional, funny, casual) Add filtering logic to skip low-value articles Connect Slack, Airtable, or Notion for internal team usage Change Gmail to SendGrid or Outlook if preferred Final Notes This workflow uses: Dumpling AI** /crawl endpoint to extract article content Split, **Code, and Edit Fields nodes to format multi-article input GPT-4o** for summarization and HTML formatting Gmail** for delivery This setup eliminates manual steps and delivers fast, consistent newsletters powered by AI.

This workflow automates web scraping of Amazon search result pages by retrieving raw HTML, cleaning it to retain only the relevant product elements, and then using an LLM to extract structured product data (name, description, rating, reviews, and price), before saving the results back to Google Sheets. It integrates Google Sheets to supply and collect URLs, BrightData to fetch page HTML, a custom n8n Function node to sanitize the HTML, LangChain (OpenRouter GPT-4) to parse product details, and Google Sheets again to store the output. URL to scape . Result Who Needs Amazon Search Result Scraping? This scraping workflow is ideal for teams and businesses that need to monitor Amazon product listings at scale: E-commerce Analysts** – Track competitor pricing, ratings, and inventory trends. Market Researchers** – Collect data on product popularity and reviews for market analysis. Data Teams** – Automate ingestion of product metadata into BI pipelines or data lakes. Affiliate Marketers** – Keep affiliate catalogs up to date with latest product details and prices. If you need reliable, structured data from Amazon search results delivered directly into your spreadsheets, this workflow saves you hours of manual copy-and-paste. Why Use This Workflow? End-to-End Automation** – From URL list to clean JSON output in Sheets. Robust HTML Cleaning** – Strips scripts, styles, unwanted tags, and noise. Accurate Structured Parsing** – Leverages GPT-4 via LangChain for reliable extraction. Scalable & Repeatable** – Processes thousands of URLs in batches. Step-by-Step: How This Workflow Scrapes Amazon Get URLs from Google Sheets – Reads a list of search result URLs. Loop Over Items – Iterates through each URL in controlled batches. Fetch Raw HTML – Uses BrightData’s Web Unlocker proxy to retrieve the page. Clean HTML – A Function node removes doctype, scripts, styles, head, comments, classes, and non-whitelisted tags, collapsing extra whitespace. Extract with LLM – Passes cleaned HTML into LangChain → GPT-4 to output JSON for each product: name, description, rating, reviews, price Save Results – Appends the JSON fields as columns back into a “results” sheet in Google Sheets. Customization: Tailor to Your Needs Adaptable Sites** – This workflow can be adapted to any e-commerce or other website, for example Walmart or eBay. Whitelist Tags** – Modify the allowedTags array in the Code node to keep additional HTML elements. Schema Changes** – Update the Structured Output Parser schema to include more fields (e.g., availability, SKU). Alternate Data Sink** – Instead of Sheets, route output to a database, CSV file, or webhook. 🔑 Prerequisites Google Sheets Credentials** – OAuth credentials configured in n8n. BrightData API token** – Stored in n8n credentials as BRIGHTDATA_TOKEN. OpenRouter API Key** – Configured for the LangChain node to call GPT-4. n8n Instance** – Self-hosted or cloud with sufficient quota for HTTP requests and LLM calls. 🚀 Installation & Setup Configure Credentials** In n8n, set up Google Sheets OAuth under “Credentials.” Add BrightData token as a new HTTP Request credential. Create an OpenRouter API key credential for the LangChain node. Import the Workflow** Copy the JSON workflow into n8n’s “Import” dialog. Map your Google Sheet IDs and GIDs to the {{WEB_SHEET_ID}}, {{TRACK_SHEET_GID}}, and {{RESULTS_SHEET_GID}} placeholders. Ensure the BRIGHTDATA_TOKEN credential is selected on the HTTP Request node. Test & Run** Add a few Amazon search URLs to your “track” sheet. Execute the workflow and verify product data appears in your “results” sheet. Tweak batch size or parser schema as needed. ⚠ Important API Rate Limits** – Monitor your BrightData and OpenRouter usage to avoid throttling. Amazon’s Terms** – Ensure your scraping complies with Amazon’s policies and legal requirements. Summary This workflow delivers a fully automated, scalable solution to extract structured product data from Amazon search pages directly into Google Sheets—streamlining your competitive analysis and data collection. 🚀 Phil | Inforeole

Idea The idea for app came since I wanted to build a unique gift for my niece because she gets excited for her birthday (which Im going to miss this year). The web app has a simple countdown (in html and JS) but more importantly, there is an AI agent that will answer some specific questions and know her preferences. How it works The questions from app are sent via web hook to N8N which has pulls preferences file (about her likes, dislikes, personality) from postgre and AI Agent that will answer questions/respond. The current status is stored back in postgre (especially about status of cat and universe happenings) before responding back. Features Integrated AI chatbot via N8N webhook Persistent conversation history Minimizable chat interface Fallback support for offline testing Features: -- Wheres Mittens - This is a query to track her lost cat in multiverse. -- Multiverse updates with recent update stored Pre Requisites Postgre SQL database is available. Alternatively, use any other database but change the N8N nodes. LLM Api Key. Step by Step Instructions Export this N8N Workflow. Modify LLM API Key, I used openAI, 4.1 For web app scofflding,you will need Node, HTML and Javascript. I've created a mini version using Node and JS with web app and N8N connection settings here: <https://github.com/productiser/FiBirthdayAgent> PostgreSQL Database Script (1 table for memory and context storage): CREATE TABLE fifi_world_context ( id TEXT PRIMARY KEY, -- e.g., 'agent_fifi' cat_location TEXT, -- e.g., "Bubble Nebula" cat_activity TEXT, -- e.g., "Playing laser tag with moon mice" fifi_preferences JSONB, -- e.g., likes/dislikes/foods/shows world_history TEXT, -- Summary of narrative events last_updated TIMESTAMP DEFAULT CURRENT_TIMESTAMP ); 5.Modify system prompt as per your needs. Built With N8N Self hosted Self hosted web app Hosted on Vercel Total spend = <£1 (AI costs only) Total Time = <1 day Support Watch this video for web app overview and how it looks. <https://youtu.be/e7PlrTdvwoM> Contact me on info@pankstr.com/ superllmuser@gmail.com for any queries Hope you enjoy!!

Building agentic AI workflows often requires multiple moving parts: memory management, document retrieval, vector similarity, and orchestration. Until now, these pieces had to be custom-wired. But with the new native n8n nodes for MongoDB Atlas, we reduce that overhead dramatically. With just a few clicks: Store and recall long-term memory from MongoDB Query vector embeddings stored in Atlas Vector Search Use these results in your LLM chains and automation logic In this example we present an ingestion and AI Agent flows that focus around Travel Planning. The different interest points that we want the agent to know about can be ingested into the vector store. The AI Agent will use the vector store tool to get relevant context about those points of interest if it needs to. Prerequisites MongoDB Atlas project and Cluster OpenAI Valid API Key for embeddings (can be other provider) Gemini API Key for the LLM (can be other provider) How it works: There are 2 main flows. One is ingesting flow: Gets a document from a webhook and use MongoDB Vector Atlas to embed the document title and description into points_of_interest collection. Embeddings are stored in a field named embedding Embeddings used are OpenAI's but it can be any type of supported embedders. Second flow is an AI Agent node with Chat Memory Stored in MongoDB Atlas and a Vector Search node as a tool: Chat Message Trigger**: Chatting with the AI Agent will trigger the conversation store in the MongoDB Chat Memory node. When data is necessary like a location search or details it will go to the "Vector Search" tool. Vector Search Tool** - uses Atlas Vector Search index created on the points_of_interest collection: // index name : "vector_index" // If you change an embedding provider make sure the numDimensions correspond to the model. { "fields": [ { "type": "vector", "path": "embedding", "numDimensions": 1536, "similarity": "cosine" } ] } Additional Resources MongoDB Atlas Vector Search n8n Atlas Vector Search docs

This workflow is an experiment to integrate charts in AI Agents, using the new Structured Output from OpenAI and Quickchart.io. How it works Users chat with an AI Agent. Anytime the AI Agent considers a chart is needed, it calls a tool to generate a chart OpenAI generates a chart using the Quickchart definition This object is added at the end of a Quickchart.io URL (see documentation) The url is added in the conversation via the AI Agent as markdown. Set up steps Create an OpenAI API Key Create the OpenAI credentials Use the credentials for the HTTP Request node (as Predefined Credential type) Activate your workflow Start chatting For example, you can ask the AI Agent to generate a chart about the top 5 movies at the box office Start exploring the limits Shout-out Quickchart.io is an amazing open source project that provides a free API to test. Go check them out! Example of chart

This workflow contains community nodes that are only compatible with the self-hosted version of n8n. Overview This workflow automates the process of transcribing audio files and summarizing them using OpenAI models, with the final output stored neatly in Notion. Whether you're a researcher, content creator, student, or professional, this automation saves time by converting voice recordings into actionable summaries with zero manual effort. Created by: Abdullah Dilshad Contact: iamabdullahdilshad@gmail.com Who It’s For This template is ideal for: Researchers**: Transcribe and summarize interviews, lectures, or research recordings. Content Creators**: Convert podcasts or videos into transcripts and social captions/show notes. Students**: Automatically turn lectures or study group audio into summarized notes. Professionals**: Log meeting notes and summaries directly into your Notion workspace. How It Works This four-step workflow performs the following: Step 1:* *Trigger: New Audio in Google Drive** Automatically triggers when a new audio file (MP3/WAV) is uploaded to a specified Google Drive folder.The file is then downloaded for processing. Step 2: Transcribe Audio with Whisper** The audio file is sent to OpenAI’s Whisper model for high-accuracy transcription. Step 3: Summarize Transcript with GPT-4** The transcript is passed to GPT-4, which generates a clean, concise summary. Step 4: Store Summary in Notion** A new Notion page is created with the generated summary and optional metadata (file name, upload time, etc.). Setup Instructions Step 1: Google Drive Trigger** Connect your Google Drive account. Select the folder you want to monitor. This node detects new file uploads and passes the file for download. Step 2: Download File** Downloads the new audio file for transcription. Step 3: Transcribe Recording (OpenAI Whisper) Connect your OpenAI API Key. Ensure this node receives the binary audio file. It will return the transcription as plain text. Step 3: Transcribe Recording (OpenAI Whisper)** Connect your OpenAI API Key. Ensure this node receives the binary audio file. It will return the transcription as plain text. Step 4: Summarize Transcript (GPT-4 via AI Agent)** Use your OpenAI API Key. Configure a summarization prompt like: "Summarize the following transcript in a clear and concise manner:" Connect the output from Whisper into this GPT-4 prompt. Step 5: Notion Integration** Connect your Notion account. Choose or create a database to store summaries. Map the GPT output (summary) to a "Text" or "Rich Text" property. Optionally include metadata like filename, file upload date, etc.

👤 Who is this for? This workflow is ideal for social media managers, personal brand strategists, ghostwriters, and founders who want to post regularly on LinkedIn without spending hours writing from scratch. It’s also useful for marketing agencies and assistants looking to automate consistent post creation using curated articles as source material. 🧩 What problem does this workflow solve? Manually reading multiple articles, extracting key insights, and writing a clean, professional LinkedIn post is a time-consuming process. This workflow automates everything: from pulling topics, finding related articles, summarizing them using AI, and even generating a matching image to accompany the post. It ensures faster content turnaround, more consistency, and less manual effort. 🔁 What this workflow does This workflow starts manually and retrieves one topic marked as “To do” from a Google Sheet. That topic is used as a search term for Dumpling AI’s search endpoint, which scrapes and returns the top three article contents related to the topic. These articles are sent to a LangChain agent powered by GPT-4o, which analyzes and summarizes the content into a LinkedIn post in a friendly, insightful tone. It also generates an image prompt for the post. After generating the post and image prompt, the data is extracted using a Set node. The prompt is sent to Dumpling AI’s image generation endpoint, which returns an image URL. Finally, the post text, image prompt, image URL, and status update (“created”) are saved back to the original row in Google Sheets. 🛠️ Workflow Breakdown Manual Trigger – Starts the automation. Google Sheets (Get Topic) – Searches for the first row in your content pipeline sheet where the “status” is “To do”. HTTP Request (Dumpling AI Search) – Uses the topic as a search query to pull 3 article contents using Dumpling AI’s API. Set LangChain GPT Model – Defines GPT-4o as the LLM for the LangChain Agent. LangChain Agent (Summarize & Generate) – Summarizes all 3 articles and generates a LinkedIn post and a related image prompt. Set (Extract Data) – Extracts postText and imagePrompt from the LangChain agent output. HTTP Request (Dumpling Image Gen) – Sends imagePrompt to Dumpling AI’s image generation endpoint. Update Google Sheets – Writes the post, image prompt, and image URL back to the sheet and changes the row status to “created”. ⚙️ Setup Instructions Dumpling AI Sign up at Dumpling AI Get your API key and connect it in the HTTP Request nodes (Search and Image endpoints) Use the /search endpoint to retrieve article content Use the /generate-image endpoint to create the image Google Sheets Create a spreadsheet with columns: topic, status, postText, imagePrompt, imageURL Add sample topics and set their status to To do LangChain (GPT-4o) Connect your OpenAI credentials to n8n Make sure GPT-4o is available in your OpenAI account Use the LangChain node to process multi-input summarization and generate a social media caption Customize the Prompt (Optional) Adjust the Set node to tweak the input format sent to the LangChain agent Add constraints like tone, hashtags, or emojis to fit your brand style 🧠 How to Customize This Workflow Change the content source (RSS feed, Notion DB, etc.) instead of Google Sheets Add a scheduler node to run this automatically every morning or weekly Use Airtable instead of Google Sheets for more control and filtering Send the final post to LinkedIn using the Buffer or LinkedIn API Add a Telegram or Slack notification when new content is ready for approval

✨ What It Does Mello is a Claude-powered Slack assistant that helps you stay on top of unread messages across all your channels. It: Summarizes conversations contextually using Claude AI. Generates reply suggestions and sends them as private (ephemeral) Slack messages. Lets you respond instantly with one-click AI-suggested replies. Perfect for busy teams, founders, and anyone looking to reduce Slack noise and save hours each week. 🔧 Setup Instructions Create a Slack App Go to Slack API → Your Apps Click Create New App and set it up for your workspace Under OAuth & Permissions, add: Bot Token Scopes: commands, chat:write, channels:history, users:read User Token Scopes: channels:history, chat:write Enable Interactivity, and point the Request URL to your n8n webhook (e.g. /slash-summarize) Add Claude API Get an API key from Claude (Anthropic) In n8n, set up the Claude API credential (or switch to OpenAI) Import This Workflow Go to your n8n instance, click Import, and paste this template Update any placeholders (Slack app, Claude key, webhook URLs) Follow the inline sticky notes for guidance Test It Type /summarize in any Slack channel Mello will fetch unread messages, summarize them, and show reply buttons in a private message ⏱ Setup time: ~10 minutes 🛠 Workflow Highlights Slash command trigger (/summarize) Slack API integration to fetch messages Claude AI for contextual summaries Reply suggestions with smart buttons Private Slack delivery (ephemeral messages) Designed to be easily extended (e.g. add support for OpenAI, custom storage) 🔒 Note This is a lite preview of the full Mello workflow. ✅ The full version includes: Slack reply buttons with thread context Full OAuth flow with token storage MongoDB integration Custom Claude/OpenAI configuration Hosted version with onboarding, branding & support 💡 Want access to the complete version? 📩 Email nina@baloon.dev

Overview This release introduces a Voice-Enabled Tour Recommendation System that leverages n8n, ElevenLabs Voice Agent, OpenAI GPT-4o, and Pinecone Vector DB to deliver personalized travel itineraries based on spoken input. Users speak their preferences to the ElevenLabs voice agent, which then triggers an n8n workflow that returns a tailored tour plan. Features Voice interaction with AI-powered travel agent via ElevenLabs Uses ChatGPT-4o for contextual understanding and generation Dynamic query handling with vector-based search using Pinecone Fast response generation using n8n webhook Modular agent memory and role design for scalable enhancement Pre-requisites n8n account with workflow creation access ElevenLabs account with agent and webhook setup OpenAI API key (GPT-4o access) Pinecone account for vector database A list of vectorized tour packages using this n8n embedder (https://creators.n8n.io/workflows/5085) Setup Instructions Step 1: Configure the Voice Agent Webhook in ElevenLabs Use POST method Webhook URL: https://... Breakdown voice input into: Destination Type of tour Number of days Number of passengers Step 2: Set Up the AI Agent Prompt in ElevenLabs Use a conversational style with summaries, clarifying questions, and affirmations. Example Prompt: “You use a natural speech style and periodically summarize... Your goal is to help callers create a personalized tour plan.” Step 3: Select LLM LLM: GPT-4o Mini Memory window: Up to 5 contexts Step 4: Integrate Tools Use Custom Tool: n8n ID: tool_xxxxxx Tool Description: “Generates travel plan once the details are collected” Step 5: Build n8n Workflow Trigger: Webhook (POST) Process user input: Tour Recommendation AI Agent Use OpenAI Chat Model (GPT-4o) for reasoning Query Pinecone Vector Store using Tour Builder Q&A node Respond with structured Itinerary Plan via webhook response How to use: Execute the n8n workflow (the webhook waits for the voice trigger from elevenlabs) Start the Elevenlabs Voice Agent Request for a tour plan to any destination giving the details of your tour preferences. Wait for the Voice Agent to respond back with tour package suggestions after fetching the tour details from the n8n workflow. Close the conversation. | Area | Improvement | | ------------------ | ----------------------------------------------------- | | 🔉 Voice UX | Natural-sounding travel agent using ElevenLabs | | 💡 Personalization | ChatGPT-4o adapts based on travel style & preferences | | 📚 Knowledge Base | Pinecone-powered vector retrieval of real tour data | | 🔁 Reusability | Modular workflow with reusable embedding tools | | ⚙️ System Design | Separation of memory, logic, and data layers | Who is this for? Travel Agencies & DMCs Offer ultra-personalized packages based on customer queries. Let AI do the matching. Tour Package Aggregators Auto-curate and send matching packages from your catalog — no manual searching needed. Content & Marketing Teams Craft customized tour recommendations for email campaigns and newsletters. Tech-enabled Travel Startups Embed this intelligence in your workflows, CRMs, or chatbots to delight customers.