Resolve Customer Support and Improve Products

This agent operates as an autonomous Support Responder, resolving customer support inquiries while systematically identifying and documenting product improvement opportunities from recurring issues and feedback patterns. It runs a five-stage process: analyzing the support request (categorizing the ticket - bug, feature request, how-to, billing - and assessing urgency by business impact, and checking whether it's a recurring issue), researching and investigating (searching documentation and known issues, checking the internal knowledge base, and identifying root cause for technical reports), crafting a response (clear, empathetic, actionable, with step-by-step instructions and alternatives when the primary solution isn't viable), documenting patterns (logging recurring issues that indicate product gaps, feature requests with business justification, and UX pain points), and escalating when necessary (flagging cases needing engineering involvement, critical bugs or security issues, or complex technical routing).

It outputs a customer-facing support response (subject line, empathetic acknowledgment, a clear step-by-step solution, relevant documentation links, and a closing offer for further help) and an internal ticket analysis report (ticket ID, category, resolution time, root cause) paired with a product insights section flagging whether a pattern was identified, the specific improvement opportunity, its business impact, and a suggested owning team.

A priority matrix classifies every ticket: critical (service down, security breach, data loss), high (feature broken, billing issues, angry customer), medium (feature requests, general questions, minor bugs), and low (enhancement ideas, general feedback, documentation requests).

Use this agent when you need customer support tickets resolved with a documented root cause and a systematic check for whether the issue is a recurring pattern worth escalating as a product gap, not just a one-off reply.

It is not a fit for tickets that are already flagged as critical security or legal issues - the agent's own escalation triggers route those to specialists rather than attempting to resolve them directly, and it explicitly never shares customer data between different support cases.

Inputs are the support ticket text and any relevant customer or product context. Outputs are a customer-facing response with a clear solution and resource links, plus an internal ticket analysis report documenting root cause, resolution time, and any product-gap pattern identified with its business impact and suggested owner.

Support teams who want every ticket resolution to also feed a structured product-feedback loop - tracking resolution rate, customer satisfaction, and time-to-response while surfacing recurring issues, feature requests, and UX pain points as actionable, owned insights rather than letting them get lost in ticket history.