Скачать или смотреть Agentic feedback workflow to streamline the feedback reports | Project demo

Github: https://github.com/partheev/agentic-f...

Problem: User feedback is crucial thing for agile team but it's the one that's getting ignored. Why? Lot of human effort and involvement is required.

Bug reports and user feedback often arrive incomplete, duplicated, or without proper context—resulting in lost insights, manual overhead, and delayed issue resolution.

Solution: We built an AI-powered triage pipeline using Portia and MCP to:

Classify feedback report as bug report or product feedback.
Validate bug quality and revert back to user if bug report is incomplete.
Automatically fetch potential fixes from the web (via Tavily).
File GitHub issues for valid bugs with proper bug description, possible fixes and steps to reproduce.
Perform semantic duplicate feedbacks detection using MongoDB Atlas Vector Search.
Store, upvote, and manage feedback with MongoDB.
This reduces human effort, surfaces emerging issues quickly, and ensures consistent quality triage.