Multi-Stage Customer Feedback Sentiment & Routing Pipeline

The Scenario

A SaaS company wants to monitor customer support feedback. They need a pipeline that retrieves new reviews, performs sentiment analysis, extracts product features mentioned, runs a secondary prompt to suggest bug-fixes or customer success responses, and routes high-priority complaints to a specific Slack channel.

The Brief

Map out a multi-stage prompt chaining workflow using Zapier/Make and an LLM. Create two chained prompts: Prompt 1 (Sentiment classification, feature extraction, and priority scoring in strict JSON format), and Prompt 2 (Drafting a tailored customer success response or dev bug ticket depending on priority). Detail how you parse the JSON from Prompt 1 to dynamically route the data.

Deliverables

• Logic diagram showing the path of customer feedback through the workflow (including branching paths based on sentiment/priority).
• Prompt 1 and Prompt 2 templates with variables.
• The exact JSON schema returned by Prompt 1.
• The routing logic configuration (e.g. conditional rules for Slack notifications vs email drafts).
• An error-handling plan detailing what happens if the LLM fails to return valid JSON.

Submit Your Work

Your submission is graded against the rubric on the right. If you pass, you get a public Badge URL you can share on LinkedIn. There is no draft save, so work offline first and paste your finished response here.