Can i now ask GrowERP to do work for me?

Yesterday evening, lying in bed before sleep, I found myself diving into the latest trends in business software development. One thing stood out above all: AI—and more AI. Yes, Artificial Intelligence is everywhere right now, popping up in product updates, strategy meetings, and tech blogs. It’s being embedded into almost every software product out there. But let’s pause and really ask—what kind of AI are we talking about?

In most applications, AI is still just a thin wrapper around a large language model like ChatGPT, Bard, or Claude. These integrations are typically limited to a simple chat interface, where users can type in questions and get back answers—or at best, interact with a few pre-configured prompts that copy and paste visible data into the AI for you. While that might seem impressive on the surface, it’s barely scratching the surface of what’s possible.

The big limitation? The AI doesn’t know your application. It doesn’t understand your data, your workflows, or your logic. It's like giving someone a microphone to talk to your system, but not giving them access to any of the backstage controls.

But that’s about to change with the arrival of MCP—the Model Context Protocol.

MCP introduces a powerful new paradigm. Instead of AI just talking about your app, you can now build an MCP server on top of your application that exposes its logic, structure, and actions. The result? AI can now ask GrowERP to do actual work.

Here’s the exciting part: there’s already an MCP server implementation that understands the OpenAPI protocol, which GrowERP’s backend already uses. This means GrowERP is already compatible—without any major overhaul. We’re standing at the edge of a massive leap forward in user interaction.

Imagine telling GrowERP:
“Generate an invoice for customer X with the last 3 shipped orders,”
or
“Create a new product listing and assign it to the Electronics category.”

And it just does it.

No need to navigate menus, tabs, or forms. Just a prompt. Just natural language.

This is the promise of a human interface to enterprise software—one where AI acts as an intelligent, contextual bridge between users and the system.

But here’s a critical question: are users ready for this shift?

Moving from traditional, menu-based screens to a conversational, prompt-driven interface is a radical change. And with that shift comes a new challenge: prompting accurately. If you've done any application development, you know how hard it is to get users to express their needs clearly. Will they learn to “prompt” well enough for this to work?

Time will tell—but the potential is huge.

The future isn’t just coming. It’s already here.

Let me know what you think!