Ask questions about past meetings using voice with OpenAI and Pinecone

📊 Description This workflow allows users to ask questions about past meetings using their voice. It converts the voice question into text, searches stored meeting notes using Pinecone, and replies with a spoken answer generated by AI. It helps teams quickly recall decisions, tasks, and discussions without reading long meeting notes. 🧠🔊

🔁 What This Template Does 1️⃣ Receives a voice question through a webhook endpoint. 🌐 2️⃣ Converts the audio question into text using speech transcription. 🎤➡️📝 3️⃣ Cleans and prepares the question for searching. ✅ 4️⃣ Converts the question into an embedding for semantic search. 🔍 5️⃣ Searches relevant meeting notes from Pinecone using the team namespace. 📚 6️⃣ Combines retrieved meeting context into a single readable format. 🧩 7️⃣ Uses AI to answer the question strictly from the meeting context. 🤖 8️⃣ Converts the AI’s text answer into spoken audio. 🔊 9️⃣ Sends the audio response back to the user via webhook. 🔁

⭐ Key Benefits ✅ Allows hands-free access to meeting information ✅ Saves time searching through meeting notes ✅ Prevents AI hallucinations using RAG ✅ Supports multiple teams using namespaces ✅ Works with voice-based tools and assistants ✅ Improves meeting recall and clarity

🧩 Features - Voice-based question input - Speech-to-text transcription - Semantic search using Pinecone - Team-based data isolation - Context-aware AI responses - Text-to-speech output - Webhook-driven architecture

🔐 Requirements - OpenAI API key for transcription, embeddings, and speech generation - Azure OpenAI credentials for chat responses - Pinecone API key with a configured index - Matching embedding model for ingest and query - Webhook client capable of sending audio files

🎯 Target Audience - Teams that conduct frequent meetings - Managers needing quick decision recall - Remote and distributed teams - Product, engineering, and operations teams - Automation builders using n8n - Organizations adopting voice-based workflows