Quick overview
This workflow runs every four hours to search public X mentions via Apify, uses an OpenAI chat model to classify each mention as a question, complaint, or noise, then routes questions to Notion, complaints to Slack, and everything else to a Google Sheets log.
How it works
- Runs every four hours on a schedule.
- Searches X for your configured search terms using the Apify Twitter Search Scraper actor and collects up to the defined number of results.
- Filters out very short mentions and skips posts already seen in previous workflow runs.
- Sends each remaining mention to an OpenAI chat model to return JSON with a route (question/complaint/noise), a one-sentence summary, and an urgency score.
- Creates a new page in a Notion database for routed questions, including the mention link.
- Posts routed complaints to a Slack channel with the urgency score, summary, and link.
- Appends all routed noise items to a Google Sheets sheet for simple logging.
Setup
- Add an Apify credential and ensure you have access to the
dami_studio~twitter-search-scraper actor.
- Add an OpenAI credential for the chat model used to classify mentions.
- Update the configuration values (searchTerms, product description, maxItems, and Slack channel) in the “What to watch for” step.
- Add Notion credentials, select the target database ID, and ensure it has a URL property for saving the mention link.
- Add Slack credentials and confirm the target channel exists and matches the configured channel value.
- Add Google Sheets credentials and select the spreadsheet and sheet to append logged items.