There is so much noise surrounding AI right now. Every day seems to bring a new tool, a new panic, or a new hype cycle. It is easy to get lost in the abstraction of it all.

So, instead of worrying about the future let's focus on present utility.

For me, the most effective way to deploy AI right now is to stop thinking of it as 'software' and start thinking of it as 'staff'.

I have built a highly efficient, multi-skilled team using tools that are available today. These tools allow me to multiply my capabilities, access skills (some of which I do not personally possess), and work with the efficiency of a much larger team.

Here is how I structure my digital workforce. This is the AI team I rely on…

The 'Confidant' & Non-Executive Director: Claude

Every leader needs a trusted advisor. For me, that role is filled by Claude.

I treat Claude as a critical friend and a personal sounding board. It is less about generating bulk content and more about high-level strategy. If I need a 'thought leader' to challenge my assumptions or help refine a complex idea, Claude is the ideal candidate. It (he) acts as the strategic partner in the room, helping to clarify my thinking before any actual work begins.

The Executive Assistant: Lindy

Strategic thinking is useless if the logistics are a mess. Enter Lindy.ai, my Executive Assistant and Marketing Co-ordinator.

Lindy handles the friction that usually slows me down. She talks to prospects, books meetings, and automatically handles my meeting preparation. But her role goes deeper than just calendar management. Lindy takes meeting notes and handles RAG-enabled tasks (Retrieval-Augmented Generation).

She even acts as a content filter; I can task Lindy with watching YouTube content and reporting back to me with summaries of actionable points. It is like having someone read the internet for me and only telling me the parts that matter.

The Head of Product & Design: Google AI Studio

When it is time to build, I turn to Google AI Studio.

Think of this tool as my Head of Product. It handles design work, prototyping, and the creation of MVP (Minimum Viable Product) apps. It is capable of taking an idea from zero to one, building code that is suitable for actual deployment on platforms like GitHub or GCP.

Crucially, it plays well with others. It works with other AI tools on my behalf, ensuring that the product vision is executed technically.

The Lead Engineer: Manus

If Google AI Studio is the product visionary, Manus is the Lead Engineer and Senior Developer who gets the difficult hands-on coding done.

Manus is characterised by high energy and a 'next-level' proactive attitude. It does not just wait for prompts; it actively pushes the development forward. Like my Head of Product, Manus works autonomously with other AI agents to execute complex technical tasks. It provides the engineering muscle required to turn a prototype into a working reality. Manus always provides documentation guides and implementation checklists. Or, can even access and update code directly.

The Editor in Chief: Gemini (Gems)

Content quality control is handled by Gemini, specifically using the "Gems" feature (a Gem is much like a custom GPT).

I view Gemini as my Editor in Chief. By defining specific personas with robust 'system prompts' (which establish the rules of engagement and long-term context), I can ensure a consistent tone of voice across different projects. Any time I ask the editorial Gem to help with a writing task, it already knows the business, its role, the preferred approach and the requisite character.

The Head of Research: Notebook LM

Finally, no team is complete without a dedicated researcher.Someone who can go deep on a task, but summarise the outputs succinctly. NotebookLM is arguably the most underrated tool on my list.

It serves as my Head of Research. It is a focused and controllable tool that allows me to point it at specific sources, ensuring that the output is grounded in facts I trust, rather than general internet noise. It can even generate a podcast summarising or discussing its findings, which is a fantastic way for me to digest large amounts of research while away from my desk.

The Built-in Support

It is also worth mentioning that many platforms now come with their own built-in AI team members, such as Omni for Airtable. These wait in-situ, ready to assist the core team with platform-specific tasks. I will often get one of the others (e.g. Claude or Gemini) to write a prompt for Airtable or Lindy, optimised for the task at hand, the particular context and for the applicable word/character limit.

Summary

You may not need to hire a dozen people to do the work of a dozen people. By assigning specific roles to specific AI models (strategy to Claude, engineering to Manus, research to NotebookLM, etc), I have built a team that liberates me from the grind.

The tools are there, ready to be guided and orchestrated by you.