Meeting Assistant: Real Time Notes, Action Items, and Insights

This workflow automates manual meeting note-taking, transcription, and real-time assistance, replacing hours of post-meeting reviews with AI-driven efficiency. It uses Webhook (n8n-nodes-base.webhook) to initiate bot creation via HTTP Request (n8n-nodes-base.httpRequest) with Recall.ai, validated by IF nodes for URL and status checks. Another HTTP Request sets up OpenAI threads, stored in PostgreSQL via Postgres node (n8n-nodes-base.postgres). The transcription branch uses a second Webhook for real-time data, logs via Postgres nodes, detects keywords with Code node (n8n-nodes-base.code), and triggers responses through OpenAI node (@n8n/n8n-nodes-langchain.openAi). This benefits sales teams, project managers, and remote workers handling 10-50 meetings monthly, ensuring accurate notes and action items without manual effort, boosting productivity.\n\nThis workflow saves 5-10 hours weekly per team member on documentation, ideal for consulting, tech, and education sectors in mid-sized firms (50-500 employees). Requires Recall.ai ($0.70/hour recording + $0.15/hour transcription), OpenAI (token-based, ~$0.03/1k tokens for GPT-4o mini), PostgreSQL (free self-hosted), n8n (free or $20/month cloud). Scalable to hundreds of meetings, limited by Recall.ai’s 500-hour/month cap and OpenAI rate limits.\n\nThis workflow requires n8n from n8n.io (self-hosted) or cloud.n8n.io. For Recall.ai: Sign up at recall.ai, get API key, add as HTTP header auth in n8n. OpenAI: Create key at platform.openai.com, add to OpenAI credentials; set assistant ID in OpenAI node. PostgreSQL: Configure connection with host, database, user, password. Webhooks: Generate URLs in n8n editor, insert transcription destination in HTTP Request’s JSON body. Configure IF nodes for validations, Code node for keywords.\n\nThis workflow is tested by POSTing sample meeting URLs to setup webhook, verifying bot join and DB entries; simulate transcripts via transcription webhook. Common errors: Invalid API keys (regenerate/check), rate limits (monitor usage/upgrade plans). Deploy by activating in n8n. Monitor logs for issues like low-confidence transcripts; optimize by refining keywords and adding error notifications. Update APIs quarterly for compatibility.", "businessValue": "Saves 5-10 hours/week on meeting notes for teams of 10+", "setupTime": "60-90 minutes", "difficulty": "Advanced", "requirements": ["Recall.ai API ($0.70/hour recording + optional transcription)", "OpenAI API key (pay-per-use token pricing)", "PostgreSQL database access (free tiers available)", "n8n installed with basic database knowledge"], "useCase": "Automating real-time AI assistance in video meetings for remote teams"