Amol Walunj

We talk a lot about AI automating work, but setting up AI products still requires a surprising amount of manual effort.

Take an AI customer support agent. You need to connect integrations, configure instructions, define business rules, set permissions, test different scenarios, and keep everything updated.

I have been thinking about how we decide whether an AI agent is actually ready for production.

Imagine you build an agent that successfully handles 98 out of 100 customer requests. On paper, that sounds impressive. Most teams would probably consider it a success.

The more AI customer support projects I work on, the more I feel support is only the starting point. Most teams begin with fairly obvious use cases: answering FAQs, handling tickets, updating orders, or triggering refunds.

That makes sense. These problems are easy to define and the value is easy to measure.

But once the AI is connected to customer data, CRM, billing, product usage, and other systems, its role can become much broader.