What AI Actually Changed For My Business

I nearly didn't write this article. When I published What Six Months of AI Taught Me back in May, the conversation around AI felt very different.

There was excitement, optimism and, particularly amongst business owners, a sense of overwhelm about what all this new technology meant and where to begin.

The overwhelm hasn't disappeared; what has changed is the backlash.

Some of it, in my opinion, is entirely justified. There are serious ethical, environmental and economic questions that deserve proper attention and governance. Other parts feel caught up in an increasingly dramatic narrative about the future of humanity.

I'm not going to tackle all of that here.

But after numerous conversations with founders and business owners, I think there's value in sharing what actually happens when you implement AI into a business.

From Experimenting with AI to Actually Living With It

When I wrote my previous article, we were nearing the end of an ambitious project at Chiron Hotel Consulting and Chiron Lifestyle Consulting: building a completely bespoke, AI-powered workflow across the business.

It took around around eight months, involved two members of my team working closely with me and required input from everyone across the business.

We had to rethink processes, connect different programmes and, perhaps most importantly, get everyone comfortable with a new way of working.

Understandably, there was some hesitation. AI and job security are so closely linked in people's minds that introducing new technology can create concerns.

By the beginning of September, however, the system was fully operational. So far, the two months since have been among the calmest, most productive and genuinely enjoyable I've experienced since starting the business.

We're working ahead of schedule. Information is organised. Things aren't being forgotten. There's much less feeling constantly busy and/or behind.

Both my team and I feel more in control of our work, with more time for the work we enjoy, rather than the admin around it. 

For me, that's where the real opportunity of AI lies.

What We Actually Built

I won't go into every technical detail, but at the heart of our system is a custom-built, dynamic knowledge database powered by Anthropic, alongside two AI agents.

We've connected our main business tools: Slack, our CRM and Canva - where we create the majority of our content - with our AI knowledge system.

Rather than having information scattered across platforms, we've created a structure in which they automatically communicate with one another and update each other.

Client meetings are recorded and summarised through AI note-taking (handled carefully for client confidentiality). Feedback and instructions received through different channels are automatically incorporated into the system, making that information available to support future work.

Essentially, we've built a continuously improving institutional memory for the business.

No matter how organised or conscientious someone is, they cannot remember every client instruction, every piece of feedback, preference and decision made on Slack. Certainly not across multiple clients, projects and months of work.

Machines, however, are extraordinarily good at collecting, organising and retrieving information. So why not let them do what they’re best at?

The result isn't that AI is suddenly doing all our marketing. It's that the people responsible for our marketing have much better information at their fingertips, and more time in which to do their work.

I want AI to make the human side of our business stronger, not replace humans.

We Haven't Replaces Anyone, We've Hired More People

There's an assumption in much of the AI conversation that efficiency must mean reducing headcount.

That hasn't been my experience. I haven't let a single person go because of AI. In fact, I've hired more people with specific skill sets.

I've also been able to develop an entirely new business, something I wouldn't have had the capacity to manage in quite the same way before.

The objective was never to replace my team, who I love working with. It was to remove as much of the repetitive, administrative and frankly boring, work as possible.

Strategy. Creativity. Relationships. Problem-solving. Judgement; these are the things I want my team to be spending their time on.

I don't believe AI can do everything, and I certainly don't believe it should. There are things we should actively choose to protect from automation, even when the technology becomes capable of doing them.

The question isn’t whether AI can do something. It should be whether it should do something.

Knowing When To Stop

One of the less-discussed problems with AI is the constant pressure to keep improving.

There's always another tool, another update, another integration and another person on LinkedIn explaining why the technology you adopted three weeks ago is already obsolete. This is exhausting.

After eight months developing our system, I made a decision. We aren't changing anything significant until after Christmas. We’re going to use what we've built, settle into our new processes and actually enjoy the benefits.

Of course, we still maintain and refine the system where necessary. But we're no longer endlessly rebuilding it. This has been very liberating.

Technology is supposed to serve the business, not become the business. So, at some point, you have to stop experimenting and start benefiting.

Two Tools Worth Exploring

1. Viktor

Viktor is an AI agent that operates within Slack and can connect with your email and other business applications. Unlike a traditional chatbot, it's designed to carry out actual work, interacting with the programmes you already use.

I use Viktor personally at founder level, rather than rolling it out across the whole team because it’s relatively expensive and credit based.

It has been one of the most useful additions to my working life, particularly for organising information, managing tasks and getting things done.

2. Lovable

Lovable allows people to build simple websites, applications and digital products by describing what they want, without needing traditional software developers. In my view it’s great for simple projects that don’t require sophisticated design and development.

Earlier this year, I launched Paris Book Club, a new membership community in Paris, using a custom-built platform designed by me but created solely with Lovable. I run the business, but I’m supported by the technology I've built around it.

Something that previously might have required a substantial development budget and specialist technical team is now within reach of a founder with an idea.

And that's one of the reasons I think the AI conversation needs to extend beyond which jobs might disappear. We also need to talk about what people can now create, that they wouldn't have otherwise.

Businesses that might never have existed, ideas that might never have made it beyond a notebook, entirely new products, services and industries. I find that really exciting.

For me Lovable doesn’t replace graphic designers and developers, it brings to life a project that would never have been built without AI.

Lovable is also an example of where we can discuss with clients what merits professional design support and development (still plenty of things!) and what can be done alone.

A More Balanced Conversation

One thing I've noticed recently is how many women have approached me with questions about AI.

I think there's a genuine gender gap in the conversation. Much of the public discussion, particularly around technical applications, is still dominated by men. 

And while there are brilliant women working in the industry, I suspect some female founders don't always see themselves represented in the people explaining how and why this technology should be used.

I'm presenting to a group of female founders shortly about our experience, and it's a conversation I'm increasingly interested in contributing to.

I think we need to make the technology more accessible, less intimidating and more relevant to the realities of running a business.

Where I've Landed

In May, I wrote that most founder-run businesses don't have an AI problem, they have a systems problem. I still believe that.

But after several months of actually working with the systems we've built, I'd add something else.

The real opportunity isn't simply to make businesses more efficient. It's to make them better places to work.

That’s always my objective. I wanted a business where we could do great work for our clients without constantly feeling overwhelmed. Where the team could feel organised and have enough mental space to be thoughtful and creative.

Where I could enjoy running the business without spending my days managing endless small tasks, and where there was room for an interesting life outside of work.

Two months into our new way of working, that's what we've achieved.

We're also now in a position to take on a couple of new clients, which feels like a nice outcome after spending much of this year concentrating on our internal operations. Message me at ellie@chironhotelconsulting.com to discuss more if that's you.

For me, the most exciting thing about AI is the possibility of giving the people I work with more freedom to do what they do best. 

And after the last few months, I'm more convinced of that than ever.

Ellie x