Assess churn risk from HubSpot tickets with DeepSeek and Slack alerts

Description

Quick overview
This workflow runs daily to pull HubSpot support tickets, aggregates them per contact, uses a DeepSeek AI Agent to assess churn risk, writes the risk level and summary back to HubSpot, and alerts Customer Success in Slack while creating HubSpot follow-up tasks for high-risk accounts.

How it works
Runs every day at 07:00 on a schedule.
Fetches all HubSpot tickets and collects key fields like subject, priority, create date, and associated contact/company identifiers.
Aggregates the tickets by contact email to compute ticket volume, urgent/high-priority count, days since the most recent ticket, and a short subject summary.
Sends each contact’s ticket summary to a DeepSeek chat agent to return a structured churn-risk level, qualitative assessment, and recommended next action.
Updates the HubSpot contact with the churn risk level and summary in custom contact properties.
Posts a churn-risk alert to a Slack channel for medium- and high-risk accounts, and also creates a HubSpot follow-up task due the next day for high-risk accounts.

Setup
Connect a HubSpot credential with permission to read tickets and update contacts (and create tasks if you want the follow-up task step).
Add a DeepSeek API credential for the chat model used by the AI Agent and structured output auto-fix.
Add a Slack credential and set the target channel (for example, #customer-success).
Create HubSpot contact properties named churn_risk_level and churn_risk_summary (and ensure your tickets include the association fields used in this template, such as associated_contact_email and associated_contact_id).
Adjust the schedule timing and any ticket priority conventions (for example, how “HIGH”/“URGENT” is represented in your HubSpot ticket priority property).