Use an AI agent to screen store support replies
A merchant let Gorgias AI Agent handle support, then switched it off after it started giving wrong answers and committing the store to things it could not do.
- What they did
- Topped Toys worked with the Gorgias team for several weeks to configure its AI Agent. They gave it instructions for support tasks, including changing and cancelling orders, and it kept running for a few months. Once the trouble began, the merchant explored other support options and then deactivated the agent.
- How it went
- For a while, the agent resolved roughly 95% of customer issues, but later it broke down on every test case the merchant reviewed. It pointed shoppers to products that were not in the catalog, could not carry out order edits or cancellations, and made commitments the business could not meet.
- Tips
- Support should be part of the launch plan, not an afterthought. Before depending on an agent, ask who will handle escalations and how quickly they will respond; in this case, L2 replies took more than 36 hours.
Try it yourself with Claude
I have [number] recent customer support conversations for [store name]. Check each exchange and label any response that is irrelevant, inaccurate, or commits us to something we cannot provide, like [refunds, delivery dates, order cancellations]. Then produce: 1) a short set of instructions for the agent, 2) the policy boundaries it must follow, and 3) the topics that should always be passed to a person. Stop before any step that would send a message, delete information, process a refund, or place an order, and wait for my approval first.
Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.
1 more report of this
Get 3 agent uses like this every morning
With the day's top AI news. Free, and one click to unsubscribe.
Related agent uses
Turn product photos into a video ad draft
A commenter used an ad agent to turn product photos and audience details into a rough video ad with script, b-roll, and voiceover.
Set up a voice agent for calls and simple support
A team used a voice agent for after-hours support, FAQs, and order-status checks, then pushed structured call details into their internal tools.
Connect a voice receptionist to your CRM with MCP
They wired a voice agent to a custom MCP server so it could list, book, and cancel appointments through their CRM.
Use an agent to brief reps on new leads
Have an agent gather prospect and company context before a rep reaches out.
Use an agent to qualify and update CRM leads
Teams used agents inside their CRM to sort leads, enrich accounts, and draft follow-up, with humans still approving outreach.
Build an after-hours call handler that books visits
They set up agents to take late calls, gather details, book appointments, and spot upset callers for follow-up.