This is a workflow that tries to answer user queries using the standard GPT-4 model. If it can't answer, it sends a message to Slack to ask for human help. It prompts the user to supply an email address. This workflow is used in Advanced AI examples | Ask a human in the documentation. To use this workflow: Load it into your n8n instance. Add your credentials as prompted by the notes. Configure the Slack node to use your Slack details, or swap out Slack for a different service.
Tags
Call n8n Workflow ToolCCode
OpenAI Chat ModelSSimple MemorySlackRelated workflows
See all AI Automation→C
Summarize Zoom meetings and create follow-ups with OpenAI, ClickUp, and Outlook
Quick overview This workflow pulls the transcript and participant list from a recent Zoom meeting, uses OpenAI to generate formal meeting minutes, emails the summary via SMTP, and runs an AI agent that creates ClickUp tasks and (when details exist) schedules a follow-up meeting in Microsoft Outlook Calendar. How it works Starts when you manually run the workflow. Fetches your scheduled Zoom meetings and keeps only those that started within the last 24 hours. Requests the meeting recording files from Zoom, stops with an error if no transcript is available, then extracts the transcript download URL. Downloads the Zoom transcript file, extracts its text, and reformats it into clean, readable dialogue. Pulls the Zoom meeting participant list (including emails) and sends the transcript and participants to OpenAI to generate structured meeting minutes. Formats the minutes into an HTML email and sends the summary via SMTP to a participant email address. Uses an OpenAI-powered agent to extract action items and follow-up details from the same transcript, then creates tasks in ClickUp and creates a follow-up event in Microsoft Outlook Calendar when scheduling information is present. Setup Ad
SSlack chatbot powered by AI
This workflow offers an effective way to handle a chatbot's functionality, making use of multiple tools for information retrieval, conversation context storage, and message sending. It's a setup tailored for a Slack environment, aiming to offer an interactive, AI-driven chatbot experience. Note that to use this template, you need to be on n8n version 1.19.4 or later.
WSend weekly blended ad performance reports with Meta Ads, Google Ads, GA4 and OpenAI
Quick overview This workflow runs every Monday at 8am, pulls last-7-days performance data from the Meta Ads API, Google Ads API, and Google Analytics 4, calculates blended KPIs, uses OpenAI to generate a client-friendly insight summary, and delivers an HTML report by Gmail while posting status update to Slack. How it works Runs every Monday at 8am on a schedule trigger. Fetches the last 7 days of spend and engagement metrics from the Meta Ads Insights endpoint. Fetches the last 7 days of campaign cost, clicks, and conversions from the Google Ads API using a streamed search query. Fetches the last 7 days of conversions and revenue by default channel group from the Google Analytics 4 Data API. Combines the three datasets and calculates blended totals for spend, conversions, revenue, and ROAS. Sends the blended metrics to OpenAI to generate a 3–4 sentence plain-language weekly performance summary. Builds a simple HTML report and sends it to the client via Gmail while posting a short “report sent” summary to a Slack channel. If the workflow execution fails, posts the error message to an ops Slack channel. Setup Add credentials and replace the placeholders for your Meta Ads token and ad