This workflow automates the process of analyzing survey responses to extract meaningful insights. It imports responses from Google Sheets, vectorizes them using OpenAI embeddings, stores them in Qdrant for clustering, and applies K-means clustering to identify groups of similar answers. Each cluster is then summarized by an LLM to generate insights, including sentiment analysis. Finally, the insights are exported back to a new Google Sheet for reporting and further analysis.
Key Features
- 🔗 Google Sheets Integration – Automatically imports raw survey responses and exports final insights.
- 🤖 OpenAI Integration – Uses embeddings and LLMs for semantic understanding and summarization.
- 📦 Qdrant Vector Database – Stores survey responses as vectors for efficient clustering and retrieval.
- 📊 K-means Clustering – Groups similar responses to ensure all perspectives are considered.
- 💡 AI-Powered Insights – Generates short, actionable summaries with sentiment classification.
- 📝 Automated Insights Export – Creates a new insights sheet for each survey run.
- ⚡ Resilient Workflow – Processes one question at a time to minimize data loss on interruptions.
- 🛠️ Customizable Pipeline – Flexible structure to adapt to other survey formats or data sources.
