[Watch] AI changed everything for software teams. Here’s what happened at BaseKit.

A few years ago, AI felt like science fiction.

Now? It’s rewriting code, reviewing pull requests and generating entire websites.

But despite all the noise around AI, there’s still one big question every software company is trying to answer:

How do you use AI properly without turning your product into a gimmick?

At BaseKit, that question has shaped almost every decision around AI adoption. And during our recent Product Roadmap Webinar, our Technical Director Mark Jeffries stated that the answer starts with one simple rule:

If it doesn’t genuinely help customers, it doesn’t ship.

AI for real value, not marketing hype

A lot of AI features right now feel like box-ticking exercises. 

But at BaseKit, the approach has been much more deliberate.

As a certified B Corp, the focus isn’t on adding AI features for headlines. It’s about solving real customer problems with responsible AI implementation.

Take website onboarding. Instead of forcing users through complicated setup flows, BaseKit  generates a  five-page website from a simple business description.

Business name and short description in. Website out. Styling, imagery and copy included.

And it doesn’t stop there. Users can also:

• Generate product descriptions with AI
• Create website copy in different tones and text length
• Generate product descriptions from uploaded product images
• Speed up store setup dramatically

The key difference? These features remove friction instead of adding more complexity.

Inside the dev team: AI coding is already here

The biggest surprise? AI hasn’t just changed the product. It’s transformed how the platform itself gets built.

At BaseKit, AI coding agents are now deeply embedded into the software development workflow.

In some cases, AI can even build entire features from scratch.

But before anyone starts panicking about robots replacing developers, Mark makes one thing very clear:

Humans are still fully in charge.

Every code change is reviewed, understood and approved by real software engineers. AI can assist, but ownership always stays with the developer.

That human-first philosophy applies across the entire engineering process, including peer reviews. AI now assists code reviews too, spotting common issues and understanding patterns across the codebase. But the final approval? Always human.

The internet is now mostly bots

According to BaseKit’s operations team, humans are no longer the majority of internet traffic hitting the platform.

In fact, during heavy bot activity, less than 20% of traffic can actually come from real people.

The rest?

• AI scrapers
• Security scanners
• Malicious bots
• Automated crawlers
• AI training systems

That means modern infrastructure teams are spending huge amounts of time filtering, analysing and blocking bad traffic just to keep platforms stable and secure.

At the same time, they also need to allow the right AI tools through.

Because if someone asks ChatGPT, Google Gemini or Claude for recommendations for local businesses, BaseKit-powered sites need to appear in those results.

Welcome to the new internet.

The future of AI is quieter than people think

The most interesting thing about BaseKit’s approach to AI is that it’s not trying to make AI the centrepiece.

The goal isn’t to overwhelm users with prompts, workflows or futuristic complexity.

It’s to quietly make everything:

• Faster
• Easier
• More intuitive
• Less frustrating

Because ultimately, small businesses don’t care whether something is ‘AI’, they care whether it helps them get online, sell more and grow faster.

And that’s where AI becomes genuinely useful.

Head here to watch our webinar in full.