Five categories of AI project: do it now to pie in the sky

Everyone wants AI to take them to the moon, but the first win is usually sitting right on your desk.

That is why our BrA1n collective works with five categories of AI project: from do it now to pie in the sky.

Because many organisations want to start with the spectacular ideas, such as digital colleagues or self-steering organisations. Fine, ambitious, but the real value? That is earned in categories 2 and 3.

This is how we divide them up:

1. Do it now, do it yourself

Low threshold. Quick result. Things you can actually do yourself.

• Minutes and action points out of meeting recordings
• Summarising and prioritising customer email
• Producing quotes from templates
• Making an internal knowledge base searchable
• Generating product copy or social posts

This is simply: less searching, less cutting and pasting, less hassle. It is allowed to just make your Monday less annoying.

2. Simple integrations

Here the AI works together with your systems and data.

• A website chatbot connected to CRM and stock
• Automatic booking of invoices
• Generating email newsletters or even entire podcasts
• Categorising and routing support tickets
• Automatic translation

This is where AI starts to really touch your processes. All of it is proven technology. Done that 😉.

3. Strategic

This asks for more time, budget and change.

• Answering email automatically
• Automated onboarding
• An AI agent for a complete work process
• Autonomous handling of support tickets
• Quality control through image recognition

This is where structural value sits. But also more risk, more dependency and more pressure to change.

4. Experimental

These projects are promising, but not yet proven. We recommend starting small and testing.

• Autonomous agents across multiple systems
• Digital twins of processes
• AI-driven product development
• Hyper-personal marketing
• Voice assistants for complex conversations

Test, measure, learn, and only then scale up.

5. Pie in the sky

This is the category of inspiring, but not feasible yet.

• AI that sets the company strategy
• A fully autonomous organisation
• An AI colleague who can do everything
• Systems that invent products and take them to market themselves

Inspiring? Certainly. But usually the technology, the data, the legislation or the trust is missing.

This model helps to put AI projects in the right place.

The question is not: "Can we do something with autonomous agents?"

But: "Where are we losing time, money or quality today?"

That is where the gain starts. Start small. Think big. Scale smart.