Triage product UAT feedback with OpenAI, Notion, Slack and Gmail

Description Automatically triage Product UAT feedback with AI, deduplicate it against your existing Notion backlog, create/update the right Notion item, and close the loop with the tester (Slack or email).

This workflow standardizes incoming UAT feedback, runs AI classification (type, severity, summary, suggested title, confidence), searches Notion to prevent duplicates, and upserts the roadmap entry for product review. It then confirms receipt to the tester and returns a structured webhook response.

Context Feature requests often arrive unstructured and get lost across channels. Product teams waste time re-triaging the same ideas, creating duplicates, and manually confirming receipt.

This workflow ensures:

Faster feature request triage

Fewer duplicates in your roadmap/backlog

Consistent structure for every feedback item

Automatic tester acknowledgement

Full traceability via webhook response

Who is this for?

Product Managers running UAT or beta programs

Product Ops teams managing a roadmap backlog

Teams collecting feature requests via forms, Slack, or internal tools

Anyone who wants AI speed with clean backlog hygiene

Requirements

Webhook trigger (form / Slack / internal tool)

OpenAI account (AI triage)

Notion account (roadmap/backlog database)

Slack and/or Gmail (tester notification)

How it works

Trigger: feedback received via webhook

Normalize & Clean: standardizes fields and cleans message

AI Triage: returns structured JSON (type, severity, title, confidence…)

Notion Dedupe & Upsert: search by suggested title → update if found, else create

Closed Loop: notify tester (Slack or email) + webhook response payload

What you get

One workflow to capture and structure feature requests

Clean Notion backlog without duplicates

Automatic tester confirmation

Structured output for downstream automation

About me : I’m Yassin a Product Manager Scaling tech products with a data-driven mindset. 📬 Feel free to connect with me on Linkedin