Workflow Templates

Overview This template is designed for Amazon sellers and advertisers who want to automate their campaign performance analysis and bidding strategy. It solves the common challenge of manually reviewing Sponsored Products reports and guessing how to adjust keywords, placements, and budgets. By combining Amazon Advertising reports with OpenAI's GPT-4o, this workflow delivers real-time, personalized optimization instructions — automatically. Features 📥 Automatically downloads Sponsored Products reports from Google Drive 🧠 Uses AI to analyze campaign, keyword, placement, targeting, and budget performance 📊 Supports both .csv and .xlsx report formats 🔁 Handles multiple ASINs and scales easily across ad accounts 📧 Sends structured optimization recommendations to your inbox via Gmail 🗂 Built-in logic to normalize filenames and correctly map reports 🧹 Includes error handling and formatting cleanup for AI-ready input Requirements To use this workflow, you’ll need: An Amazon Ads account with access to Sponsored Products reports A Google Drive folder where Amazon Ads reports are delivered (manually or via Gmail automation) A Gmail account (for sending summaries) An OpenAI API key with access to GPT-4o Optional: a developer account for the Amazon Ads API to fully automate report generation in the future Setup Instructions 📂 Connect your Amazon Ads reports folder in the Google Drive node 🔐 Add your credentials to the OpenAI and Gmail nodes 📝 Schedule five reports in the Amazon Ads Console: Search Term Report → Detailed Targeting Report → Detailed Campaign Report → Summary Placement Report → Summary Budget Report → Summary Use “Last 30 Days”, “Daily”, and .xlsx or .csv format 🔁 (Optional) Automate report ingestion using Gmail + Drive workflows 🧪 Test with one account, then replicate across additional ad accounts as needed ⏱️ Setup time: 15–30 minutes 📌 All field-specific guidance is included in workflow notes`

This workflow contains community nodes that are only compatible with the self-hosted version of n8n. This workflow scrapes job listings on indeed via Apify, automatically gets that dataset, extracts information about the listing filters jobs off relevance, finds a decision maker at the company and updates a database (google sheets) with that info for outreach. All you need to do is run Apify actor then the database will update with the processed data. Benefits: Complete Job search Automation - A webhook monitors the Apify actor which sends a integration and starts the process AI-Powered Filter - Uses ChatGPT to analyze content/context, identify company goals, and filters based on job description Smart Duplicate Prevention - Automatically tracks processed job listings in a database to avoid redundancy Multi-Platform Intelligence - Combines Indeed scraping, web research via Tavily, and enriches each listing Niche Focus - Process content from multiple niches 6 currently (hardcoded) but can be changed to fit other niches (just prompt the "job filter" node) How It Works: Indeed Job Discovery: Search and apply filter for relevant job listings, copy and use URL in Apify Uses Apify's Indeed job scraper to scrape job listings from the URL of interest Automatically scrapes the information, stores it in a dataset and initiates a integration Oncoming Data Processing: Loops over 500 items (can be changed) with a batch size of 55 items (can be changed) to avoid running into API timeouts. Multiple filters to ensure all fields are scrapped with our required metrics (website must exist and number of employees < 250) Duplicate job listings are removed from oncoming batch to be processed Job Analysis & Filter: An additional filter to remove any job listing from the oncoming batch if it already exists in the google sheets database Then all new job listings gets pasted to chatGPT which uses information about the job post/description to determine if it is relevant to us All relevant jobs get a new field "verdict" which is either true or false and we keep the ones where verdict is true Enrich & Update Database: Uses Tavily to search for a decision maker (doesn't always finds one) and populate a row in google sheet with information about the job listing, the company and a decision maker at that company. Waits for 1 minute and 30 seconds to avoid google sheets and chatGPT API timeouts then loops back to the next batch to start filtering again until all job listings are processed Required Google Sheets Database Setup: Before running this workflow, create a Google Sheets database with these exact column headers: Essential Columns: jobUrl - Unique identifier for job listings title - Position Title descriptionText - Description of job listing hiringDemand/isHighVolumeHiring - Are they hiring at high volume? hiringDemand/isUrgentHire - Are they hiring at high urgency? isRemote - Is this job remote? jobType/0 - Job type: In person, Remote, Part-time, etc. companyCeo/name - CEO name collected from Tavily's search icebreaker - Column for holding custom icebreakers for each job listing (Not completed in the workflow. I will upload another that does this called "Personalized IJSFE") scrapedCeo - CEO name collected from Apify Scraper email - Email listed on for job listing companyName - Name of company that posted the job companyDescription - Description of the company that posted the job companyLinks/corporateWebsite - Website of the company that posted the job companyNumEmployees - Number of employees the company listed that they have location/country - Location of where the job is to take place salary/salaryText - Salary on job listing Setup Instructions: Create a new Google Sheet with these column headers in the first row Name the sheet whatever you please Connect your Google Sheets OAuth credentials in n8n Update the document ID in the workflow nodes The merge logic relies on the id column to prevent duplicate processing, so this structure is essential for the workflow to function correctly. Feel free to reach out for additional help or clarification at my gmail: terflix45@gmail.com and I'll get back to you as soon as I can. Set Up Steps: Configure Apify Integration: Sign up for an Apify account and obtain API key Get indeed job scraper actor and use Apify's integration to send a HTTP request to your n8n webhook (if test URL doesn't work use production URL) Use Apify node with Resource: Dataset, Operation: Get items and use your Api key as your credentials Set Up AI Services: Add OpenAI API credentials for job filtering Add Tavily API credentials for company research Set up appropriate rate limiting for cost control Database Configuration: Create Google Sheets database with provided column structure Connect Google Sheets OAuth credentials Configure the merge logic for duplicate detection Content Filtering Setup: Customize the AI prompts for your specific niche, requirements or interest Adjust the filtering criteria to fit your needs

This workflow allows users to generate AI videos using the cheaper model Google Veo3 Fast, save them to Google Drive, generate optimized titles with GPT-4o, and automatically upload them to YouTube and TikTok with Upload-Post. The entire process is triggered from a Google Sheet that acts as the central interface for input and output. IT automates video creation, uploading, and tracking, ensuring seamless integration between Google Sheets, Google Drive, Google Veo3 Fast, TikTok and YouTube. Benefits of this Workflow 💡 No Code Interface**: Trigger and control the video production pipeline from a simple Google Sheet. ⚙️ Full Automation**: Once set up, the entire video generation and publishing process runs hands-free. 🧠 AI-Powered Creativity**: Generates engaging YouTube and TikTok titles using GPT-4o. Leverages advanced generative video AI from Google Veo3. 📁 Cloud Storage & Backup**: Stores all generated videos on Google Drive for safekeeping. 📈 YouTube Ready**: Automatically uploads to YouTube with correct metadata, saving time and boosting visibility. 📈 TikTok Ready**: Automatically uploads to TikTok with correct metadata, saving time and boosting visibility. 🧪 Scalable**: Designed to process multiple video prompts by looping through new entries in Google Sheets. 🔒 API-First**: Utilizes secure API-based communication for all services. How It Works Trigger: The workflow can be started manually ("When clicking ‘Test workflow’") or scheduled ("Schedule Trigger") to run at regular intervals (e.g., every 5 minutes). Fetch Data: The "Get new video" node retrieves unfilled video requests from a Google Sheet (rows where the "VIDEO" column is empty). Video Creation: The "Set data" node formats the prompt and duration from the Google Sheet. The "Create Video" node sends a request to the Fal.run API (Google Veo3 Fast) to generate a video based on the prompt. Status Check: The "Wait 60 sec." node pauses execution for 60 seconds. The "Get status" node checks the video generation status. If the status is "COMPLETED," the workflow proceeds; otherwise, it waits again. Video Processing: The "Get Url Video" node fetches the video URL. The "Generate title" node uses OpenAI (GPT-4.1) to create an SEO-optimized YouTube and TikTok title. The "Get File Video" node downloads the video file. Upload & Update: The "Upload Video" node saves the video to Google Drive. The "HTTP Request" node uploads the video to YouTube via the Upload-Post API. The "HTTP Request" node uploads the video to TikTok via the Upload-Post API. The "Update Youtube URL" and "Update result" nodes update the Google Sheet with the video URL and YouTube link. Set Up Steps Google Sheet Setup: Create a Google Sheet with columns: PROMPT, DURATION, VIDEO, and YOUTUBE_URL. Share the Sheet link in the "Get new video" node. API Keys: Obtain a Fal.run API key (for Veo3) and set it in the "Create Video" node (Header: Authorization: Key YOURAPIKEY). Get an Upload-Post API key (for YouTube uploads) and configure the "HTTP Request" node (Header: Authorization: Apikey YOUR_API_KEY). Get an Upload-Post API key (for TikTok uploads) and configure the "HTTP Request" node (Header: Authorization: Apikey YOUR_API_KEY). YouTube Upload Configuration: Replace YOUR_USERNAME in the "HTTP Request" node with your Upload-Post profile name. Schedule Trigger: Configure the "Schedule Trigger" node to run periodically (e.g., every 5 minutes). Need help customizing? Contact me for consulting and support or add me on Linkedin.

How It Works – Your AI Marketing Team in Action This automation acts as your AI-powered content and image marketing assistant inside Telegram. With just a voice note or text message, it can: 🧠 Understand your request – Whether you send a message or speak into Telegram, it transcribes and processes your input using GPT-4. 🎨 Create and edit content – Based on what you say, it can generate: ✍️ Blog posts 💼 LinkedIn posts 🎬 Faceless videos 🖼️ AI-generated images 🪄 Edits to existing images 🔎 Searches through your image database 💬 Replies directly in Telegram – It sends you back the result—whether that’s a post, image, or video link—without leaving the app. 🧩 Built using LangChain agent logic – It intelligently chooses the right tool from a suite of sub-workflows like "Create Image", "Blog Post", or "Video" using agent reasoning. 🛠️ Setup Steps – Get Started in Minutes! ⌛ Time Estimate: ~15–30 minutes (faster if you're familiar with n8n) 🔗 1. Import the Template Pack 📥 Download and install these workflows into your n8n: Create Image, Edit Image, Search Images Blog Post, LinkedIn Post, Video 🔐 2. Add Required Credentials Telegram Bot 🤖 OpenRouter AI 🧠 Tavily API (for smart research) 📚 ElevenLabs 🎙️ (for voice in videos) PiAPI & Runway 🎞️ (for faceless videos) 🧩 3. Link the Tools to the Agent Node – Make sure the "Marketing Team Agent" is connected to each of the content creation tools as shown in the workflow. 📎 4. Download Templates & Logs 🧾 Google Sheets Log Template (to track output) 🖼️ Creatomate Template (optional for enhanced image control – shared in Skool group) 📌 Pro Tip: All detailed step-by-step setup instructions are included as sticky notes inside the n8n canvas. Just follow along!

📊 This AI sub-agent aggregates Tesla (TSLA) trading signals across multiple timeframes using real-time technical indicators and candlestick behavior. It is a core component of the Tesla Quant Trading AI system. Powered by GPT-4.1, it consolidates 15-minute, 1-hour, and 1-day indicators, adds candlestick pattern data, and produces a unified JSON signal for downstream use by the master agent. ⚠️ This agent is not standalone. It is triggered by the Tesla Quant Trading AI Agent via Execute Workflow. 🧠 Requires: 4 connected sub-agents and Alpha Vantage Premium API Key 🔌 Required Sub-Workflows To use this workflow, you must install: Tesla 15min Indicators Tool Tesla 1hour Indicators Tool Tesla 1day Indicators Tool Tesla 1hour and 1day Klines Tool Tesla Quant Technical Indicators Webhooks Tool (provides Alpha Vantage data) 🧠 What This Agent Does Fetches pre-cleaned 20-point JSON outputs from the 4 sub-agents listed above Analyzes each timeframe individually: 15m: momentum and short-term setups 1h: confirmation of emerging trends 1d: macro positioning and trend alignment Klines: candlestick reversal patterns and volume divergence Generates a structured final signal in JSON with: Trading stance: Buy, Sell, Hold, or Cautious Confidence score (0.0–1.0) Multi-timeframe indicator breakdown Candlestick and volume divergence annotations 📋 Sample Output { "summary": "TSLA momentum is weakening short-term. 1h MACD shows bearish crossover, RSI declining. 1d candles confirm potential reversal setup.", "signal": "Cautious Sell", "confidence": 0.81, "multiTimeframeInsights": { "15m": { "RSI": 68.3, "MACD": { "macd": 0.53, "signal": 0.61 }, ... }, "1h": { "RSI": 65.0, "MACD": { "macd": -0.32, "signal": 0.11 }, ... }, "1d": { "BBANDS": { ... }, ... }, "candlestickPatterns": { "1h": "Doji", "1d": "Bearish Engulfing" }, "volumeDivergence": { "1h": "Bearish", "1d": "Neutral" } } } 🛠️ Setup Instructions Import this workflow into n8n Name it: Tesla_Financial_Market_Data_Analyst_Tool Add Required API Credentials Alpha Vantage Premium (via HTTP Query Auth) OpenAI GPT-4.1 for reasoning and synthesis Link Required Sub-Agents Connect the 4 tool workflows listed above to their respective Tool Workflow nodes Connect the webhook provider for data fetches Set Up as Sub-Agent This workflow must be triggered using Execute Workflow from the parent agent Pass in: message (optional context) sessionId (used for memory continuity) 🧾 Sticky Notes Provided 📘 Tesla Financial Market Data Analyst — Core logic overview 📈 15m / 1h / 1d Tool Notes — Indicator lists + use cases 🕯️ Klines Tool Note — Candlestick and volume divergence patterns 🧠 GPT Reasoning Note — GPT-4.1 handles final synthesis 🧩 Sub-Workflow Trigger — Proper integration with parent agent 🧠 Memory Buffer — Maintains session context across evaluations 🔒 Licensing & Support © 2025 Treasurium Capital Limited Company The logic, prompt design, and multi-agent architecture are proprietary and IP-protected. For support or collaboration inquiries: 🔗 Don Jayamaha – LinkedIn 🔗 n8n Creator Profile 🚀 Unify your Tesla trading logic across timeframes—automated, AI-powered, and built for scalers and swing traders.

🧠 Long-Term Memory System for AI Agents with Vector Database Transform your AI assistants into intelligent agents with persistent memory capabilities. This production-ready workflow implements a sophisticated long-term memory system using vector databases, enabling AI agents to remember conversations, user preferences, and contextual information across unlimited sessions. 🎯 What This Template Does This workflow creates an AI assistant that never forgets. Unlike traditional chatbots that lose context after each session, this implementation uses vector database technology to store and retrieve conversation history semantically, providing truly persistent memory for your AI agents. 🔑 Key Features Persistent Context Storage**: Automatically stores all conversations in a vector database for permanent retrieval Semantic Memory Search**: Uses advanced embedding models to find relevant past interactions based on meaning, not just keywords Intelligent Reranking**: Employs Cohere's reranking model to ensure the most relevant memories are used for context Structured Data Management**: Formats and stores conversations with metadata for optimal retrieval Scalable Architecture**: Handles unlimited conversations and users with consistent performance No Context Window Limitations**: Effectively bypasses LLM token limits through intelligent retrieval 💡 Use Cases Customer Support Bots**: Remember customer history, preferences, and previous issues Personal AI Assistants**: Maintain user preferences and conversation continuity over months or years Knowledge Management Systems**: Build accumulated knowledge bases from user interactions Educational Tutors**: Track student progress and adapt teaching based on history Enterprise Chatbots**: Maintain context across departments and long-term projects 🛠️ How It Works User Input: Receives messages through n8n's chat interface Memory Retrieval: Searches vector database for relevant past conversations Context Integration: AI agent uses retrieved memories to generate contextual responses Response Generation: Creates informed responses based on historical context Memory Storage: Stores new conversation data for future retrieval 📋 Requirements OpenAI API Key**: For embeddings and chat completions Qdrant Instance**: Cloud or self-hosted vector database Cohere API Key**: Optional, for enhanced retrieval accuracy n8n Instance**: Version 1.0+ with LangChain nodes 🚀 Quick Setup Import this workflow into your n8n instance Configure credentials for OpenAI, Qdrant, and Cohere Create a Qdrant collection named 'ltm' with 1024 dimensions Activate the workflow and start chatting! 📊 Performance Metrics Response Time**: 2-3 seconds average Memory Recall Accuracy**: 95%+ Token Usage**: 50-70% reduction compared to full context inclusion Scalability**: Tested with 100k+ stored conversations 💰 Cost Optimization Uses GPT-4o-mini for optimal cost/performance balance Implements efficient chunking strategies to minimize embedding costs Reranking can be disabled to save on Cohere API costs Average cost: ~$0.01 per conversation 📖 Learn More For a detailed explanation of the architecture and implementation details, check out the comprehensive guide: Long-Term Memory for LLMs using Vector Store - A Practical Approach with n8n and Qdrant 🤝 Support Documentation**: Full setup guide in the article above Community**: Share your experiences and get help in n8n community forums Issues**: Report bugs or request features on the workflow page Tags: #AI #LangChain #VectorDatabase #LongTermMemory #RAG #OpenAI #Qdrant #ChatBot #MemorySystem #ArtificialIntelligence

📒 Telegram Expense Tracker to Google Sheets with GPT-4.1 👤 Who’s it for This workflow is for anyone who wants to log their daily expenses by simply chatting with a Telegram bot. Ideal for: Individuals who want a quick way to track spending Freelancers who log receipts and purchases on the go Teams or small business owners who want lightweight expense capture ⚙️ How it works / What it does User sends a text message on Telegram describing an expense (e.g., “Bought coffee for 50k at Highlands”) Message format is validated If the message is text, it proceeds to GPT-4.1 Mini for processing. If it's not text (e.g. image or file), the bot sends a fallback message. OpenAI GPT-4.1 Mini parses the message and returns: relevant: true/false expense_record: structured fields (date, amount, currency, category, description, source) message: a friendly confirmation or fallback If valid: The bot replies with a fun acknowledgment The data is saved to a connected Google Sheet If invalid: A fallback message is sent to encourage proper input 🛠️ How to set up 1. Telegram Bot Setup Create a bot using BotFather on Telegram Copy the bot token and paste it into the Telegram Trigger node 2. Google Sheet Setup Create a Google Sheet with these columns: Date | Amount | Currency | Category | Description | SourceMessage Share the sheet with your n8n service account email 3. OpenAI Configuration Connect the OpenAI Chat Model node using your OpenAI API key Use GPT-4.1 Mini as the model Apply a system prompt that extracts structured JSON with: relevant, expense_record, and message 4. Add Parser Use the Structured Output Parser node to safely parse the JSON response 5. Conditional Logic Nodes Is text message? Checks if the message is in text format Supported scenario? Checks if relevant = true in the LLM response 6. Final Actions If relevant**: Send confirmation via Telegram Append row to Google Sheet If not relevant**: Send fallback message via Telegram ✅ Requirements Telegram bot token OpenAI GPT-4.1 Mini API access n8n instance (self-hosted or cloud) Google Sheet with access granted to n8n Basic understanding of n8n node configuration 🧩 How to customize the workflow | Feature | How to Customize | |----------------------------------|-------------------------------------------------------------------| | Add multi-currency support | Update system prompt to detect and extract different currencies | | Add more categories | Modify the list of categories in the system prompt | | Track multiple users | Add username or chat ID column to the Google Sheet | | Trigger alerts | Add Slack, Email, or Telegram alerts for specific expense types | | Weekly summaries | Use a cron node + Google Sheet query + Telegram message | | Visual dashboards | Connect the sheet to Looker Studio or Google Data Studio | Built with 💬 Telegram + 🧠 GPT-4.1 Mini + 📊 Google Sheets + ⚡ n8n

If you have multiple users managing workflows, there may come a time where a user “accidentally” turns off a workflow. Or, if you have workflows that automatically turn off other workflows, that code might “accidentally” turn off the wrong one. In either case, here’s a workflow that can attempt to “auto-start” accidentally disabled workflows: How it works: When activated, then every 4 hours, the workflow will search all other workflows that have the auto_resume:true tag present. If any other workflow has auto_resume:true set but is currently turned off, then this workflow will turn it back on. Of course, this watchdog won’t work if the watchdog workflow is turned off. That said, we’ve found this useful in recovering from accidental actions that cause production workflows to be turned off.

This n8n workflow automates replying to Google Play Store reviews using AI. It analyzes each review’s sentiment and tone and posts a human-like response — saving time for indie devs, founders, and PMs managing multiple apps. 💡 Use Cases Respond to reviews at scale without sounding robotic Prioritize negative sentiment feedback Maintain consistent tone and support messaging Free up time for teams to focus on product instead of ops 🧠 How it works Uses the Play Store API to fetch new app reviews Filters out reviews that have already been replied to Analyzes sentiment using OpenAI GPT-4o Passes sentiment and review context to an AI Agent node that crafts a reply Replies are posted to Play Store via Google API (Optional) Logs the reply to Slack for visibility 🛠️ Setup Instructions (Sticky notes included in the workflow) 1. HTTPS Node Replace the package name with your app’s package ID Add Google Service Account credentials → Create from Google Cloud Console with access to Play Console → Add to n8n Credential Manager 2. OpenAI Node Add your OpenAI API key → GPT-4o or GPT-4o mini supported → Customize model or instructions if needed 3. AI Agent Node Modify prompt to reflect your app name, tone, and feature set → E.g. polite, witty, casual, support-friendly, etc. → You can add reply conditions or logic for different types of reviews 4. Slack Node (Optional) Configure Slack Webhook or OAuth credentials if you want reply logs → Otherwise, delete the node to simplify the workflow ⚡ Requirements Google Play Developer Console access Google Cloud Project with service account OpenAI account (GPT-4o or mini) (Optional) Slack workspace & app for logging 🙌 Don’t want to set this up yourself? I’ll do it for you. Just drop me an email: imarunavadas@gmail.com Let’s automate the boring stuff so you can focus on growth. 🚀

:car: Business Value Proposition Accelerates ISO 26262 compliance for automotive/industrial systems by automating safety analysis while maintaining rigorous audit standards. :gear: How It Works graph TD A[Engineer uploadssystem description] --> B(LLM identifies hazards) B --> C(LLM scores risks per ISO 26262) C --> D(Generates mitigation strategies) D --> E(Produces audit-ready reports) :chart_with_upwards_trend: Key Benefits Time 50-70% faster than manual HAZOP/FMEA sessions Instant report generation vs. weeks of documentation Risk Mitigation Pre-validated templates reduce human error Auto-generated traceability simplifies audits :warning: Governance Controls Human-in-the-loop: All LLM outputs require engineer sign-off Version tracking: Full history of modifications Audit mode: Export all decision rationales :computer: Technical Requirements Runs on existing n8n instances Docker deployment (<1hr setup) Integrates with JAMA/DOORS (optional) :wrench: Setup and Usage Prerequisites Docker (Install Guide) Docker Compose (Install Guide) n8n instance (Free Self-Hosted or Cloud - Paid) OpenAI API key (Get Key) Enterprise-ready deployment: When supported by IT infrastructure teams, this solution transforms into a scalable AI safety assistant, providing real-time HARA guidance akin to engineering Co-pilot tools. :arrow_down: Installation and :play_or_pause_button: Running the Workflow For installation procedures and usage of workflow, refer the repository :warning: Validation & Limitations AI-Assisted Analysis Considerations | Advantage | Mitigation Strategy | Implementation Example | |-----------|---------------------|------------------------| | Rapid hazard identification | Human validation layer | Manual review nodes in workflow | | Consistent S/E/C scoring | Rule-based validation | ASIL-D → Redundancy check | | Edge case coverage | Cross-reference with historical data | Integration with incident databases | Critical Validation Steps AI Output Review node in n8n Example: (by code) { "type": "function", "parameters": { "functionCode": "if ($input.item.json.ASIL === 'D' && !$input.item.json.redundancy) throw new Error('ASIL D requires redundancy');" } } Version Control Prompt versions tied to ISO standard editions (e.g., ISO26262:2018-v1.2) Git-tracked changes to ai_models/training_data/ Audit trails Providing a log structure for audit trails Log structure /logs/ └── YYYY-MM-DD/ ├── hazards_approved.log └── hazards_rejected.log

Quick overview This workflow turns WhatsApp into a multimodal assistant that answers text, voice notes, images, and PDF documents using OpenAI, and can generate images with DALL·E on demand, replying back in WhatsApp with either text, audio, or an image. How it works Triggers when a new WhatsApp message is received. Detects whether the incoming message is text, a voice note, an image, a PDF document, or an unsupported type. For text messages, generates an image with OpenAI DALL·E when the message starts with the configured /image command, and sends the generated image back to the sender. For voice notes, fetches the WhatsApp media URL, downloads the audio, transcribes it with OpenAI, and forwards the transcript to the assistant. For images, fetches the WhatsApp media URL, downloads the image, describes it with OpenAI vision using the configured analysis prompt, and forwards the description (plus any caption) to the assistant. For PDF documents, downloads the file from WhatsApp, extracts the PDF text, and forwards the extracted content (plus any caption) to the assistant, while non-PDF files receive an error message. Uses an OpenAI chat model with per-contact short-term memory to generate a reply, then sends it back to WhatsApp as audio (via OpenAI text-to-speech) for voice-note conversations or as a text message otherwise. Setup Connect WhatsApp Business Cloud credentials for the WhatsApp Trigger and all WhatsApp send/media nodes, and set up the webhook callback URL in Meta so incoming messages reach n8n. Create an HTTP Header Auth credential for WhatsApp media downloads (Authorization: Bearer ) and select it in the three HTTP Request download steps for audio, images, and files. Add an OpenAI API credential with access to a chat model (gpt-4o-mini), image analysis, audio transcription, text-to-speech, and DALL·E image generation. In the Configuration step, set your WhatsApp phone number ID and optionally adjust the /image command, image size/style/quality, the image analysis prompt, and the TTS voice and voice-to-voice toggle. Additional info Run a multimodal WhatsApp AI assistant with text, voice, images, PDFs and image generation 📥 Open full documentation on Notion Need help customizing? Contact me for consulting and support : Linkedin MY NEW YOUTUBE CHANNEL 👉 Subscribe to my new YouTube channel. Here I'll share videos and Shorts with practical tutorials and FREE templates for n8n.

This workflow implements an advanced AI automation agent (OpenClaw Agent) that interacts with users through Telegram and integrates multiple AI models, external tools, and cloud services to automate complex tasks. I've described my basic idea in this video. VERY IMPORTANT: By adapting the system prompt, inserting subworkflows or mcp servers and adjusting with webhooks many of the workflows I have developed on this page it is possible to potentially extend the template infinitely. The agent can autonomously decide which tools to use to complete the request. It has access to multiple integrations, including: Gmail (read, send, reply, draft emails) Google Calendar (create, delete, check availability) Google Drive, Docs, Sheets, and Slides Web search and web scraping tools RAG (Retrieval-Augmented Generation) using a vector database Here how set up Image and video analysis tools Mathematical calculator tools Custom sub-workflows and MCP integrations The system also includes persistent chat memory stored in PostgreSQL, allowing the AI to remember previous interactions and maintain conversation context. Finally, the workflow generates a response and sends it back to the user via Telegram. If the input was voice-based, the response can also be converted into audio and returned as a voice message. An escalation mechanism allows the system to transfer the conversation to a human operator when needed. Key Advantages 1. ✅ Multimodal Interaction The workflow supports text, voice, and image inputs, allowing users to interact with the system naturally. 2. ✅ AI-Powered Automation The integrated AI agent can autonomously decide which tools to use to solve tasks, reducing manual intervention and enabling intelligent automation. 3.✅ Deep Integration with Google Workspace The workflow can manage emails, documents, spreadsheets, presentations, and calendar events directly through AI commands. 4. ✅ Retrieval-Augmented Knowledge (RAG) The system can access external knowledge sources through a vector database, improving response accuracy and enabling knowledge-based answers. 5. ✅ Persistent Memory Conversation history is stored in PostgreSQL, allowing the agent to maintain context and provide more relevant responses over time. 6. ✅ Web Intelligence Built-in web search and scraping capabilities allow the agent to gather real-time information from the internet. 7. ✅ Voice Response Capability The system can generate audio responses, creating a more natural conversational experience. 8. ✅ Modular and Scalable Architecture The workflow is highly modular, allowing new tools, agents, and services to be added easily. 9. ✅ Human Escalation When automation is not sufficient, the system can escalate the conversation to a human operator. 10. ✅ Fully Automated Digital Assistant Overall, the workflow acts as a fully autonomous AI assistant capable of performing complex operational tasks across multiple platforms. How it works This workflow is a comprehensive Telegram-based AI orchestrator that simulates an OpenClaw-style multi-agent architecture. When a user sends a message to the Telegram bot, the workflow: Receives and authorizes the message through a Telegram trigger, checking if the user ID matches an authorized user (configured in the Code node) Routes different content types using a Switch node that detects whether the incoming message contains text, voice, or images: Text messages: Directly passed to the orchestrator Voice messages: Downloaded and transcribed using OpenAI's audio transcription Images: Downloaded, uploaded to an FTP server, and transformed into a URL for processing Feeds the processed input (text, transcribed voice, or image URL with caption) into the "OpenClaw Agents" node - an AI agent configured with Gemini as the language model and Postgres for chat memory Orchestrates specialized sub-agents through the main AI agent, which can delegate tasks to multiple tools: Research and web search (Perplexity AI) Web scraping (ScrapeGraphAI) Google services (Gmail, Drive, Docs, Sheets, Slides, Calendar) RAG (Qdrant vector store with Cohere reranker) Image/Video generation (via sub-workflow) Calculator Telegram communication tools Handles response delivery based on the original message type: If the original message was voice, the workflow generates an audio response using OpenAI TTS and sends it back as audio Otherwise, sends a text response via Telegram Includes escalation capabilities through a human-in-the-loop tool for situations requiring human intervention Set up steps Configure Telegram Bot Create a Telegram bot via BotFather to get a bot token Set up Telegram credentials in n8n Replace XXX in the Code node with your authorized Telegram user ID Set up API Keys and Credentials OpenAI: API key for audio transcription and text-to-speech Google Gemini: API key for the main language model Google Cloud: OAuth2 credentials for all Google services (Gmail, Drive, Calendar, Docs, Sheets, Slides) Perplexity AI: API key for research and web search ScrapeGraphAI: API key for web scraping Cohere: API key for reranker functionality PostgreSQL: Database connection for chat memory Configure External Services Qdrant: Set up vector database for RAG functionality FTP Server: Configure FTP credentials and path (replace /XXX/ in Upload image node with actual path and domain in Set Image Url node) Configure Webhooks and Endpoints Set the Telegram webhook URL to point to your n8n instance The workflow includes multiple MCP endpoints that need proper configuration Review and Adjust Parameters Check all nodes with placeholder values (marked with XXX) Verify language settings (Italian is set for transcription) Review the system prompt in the OpenClaw Agents node and customize if needed Ensure all tool descriptions match your use case Test and Activate Run the workflow in inactive mode first to test with sample inputs Check that all connections between nodes are correct Activate the workflow when everything is verified 👉 Subscribe to my new YouTube channel. Here I’ll share videos and Shorts with practical tutorials and FREE templates for n8n. Need help customizing? Contact me for consulting and support or add me on Linkedin.