My doctor sent a prescription to a large pharmacy chain. They use virtual agents to answer the phone. I called to check if my prescription was ready and the virtual agent told me it hadn't been sent. That wasn't true. It had been sent but was rejected because of an error in the quantity.
I spent days going back and forth between my doctor and the virtual agent, getting nowhere. Finally I asked to be transferred to a human being. That person checked and gave me the real answer in two minutes. The prescription had been rejected and needed to be resubmitted with the correct quantity.
The technology was ready. What was missing was the tooling and access the human representative had. It worked fine for the happy path, but the moment something went wrong it fell apart, and nothing in the system flagged the failure.
My Analysis as an IT Insider They wanted to reduce call wait time, so they stood up a virtual agent. The ROI on this probably showed success. It cost me trust in their IT execution and produced more call volume over time. If I had bothered to take a survey, their CSAT (customer satisfaction) score would have taken a hit too.
Three things their architecture was missing:
- Authoritative system of record reads, not just the surface delivered status
- A signal for the agent to know when it doesn't have enough information to answer
- Human handoff with context, so the customer doesn't restart from scratch
My View as an IT Outsider AI is dumb and can't figure out basic things. I'm annoyed with how they wasted my time over days and I had to speak to a real person anyway.
How are you making sure your implementations are ready for the real world?
Written by Duane Grey
AI Strategy & Implementation
Independent AI consultant helping companies cut through hype and deploy systems that produce real results.