A Founder Told Me SaaS Is Dead. Here's What I Think.

A founder recently told me:

"SaaS is dead."

Recently, I had a thought-provoking conversation with a connection who's an expert in AI and automation. When I mentioned SaaS platforms, he took a bold stance: he said SaaS is dead.

That got my attention. After a few messages, he clarified:
"Not software. The SaaS pricing model."

His argument was interesting.

AI is making it easier to generate software on demand.

Instead of buying a one-size-fits-most SaaS product, businesses may increasingly expect solutions tailored to their exact workflows.

And honestly, I think there is some truth in that.

AI is definitely reducing the barrier to build.

But from an engineering perspective, building software has never been the hardest part.

Operating it is.

The questions I keep thinking about are:
* How do you guarantee consistent output?
* How do you maintain reliability across thousands of users?
* How do you control infrastructure costs?
* How do you support and evolve hundreds of customised workflows?

Those problems don't disappear with AI.

In many cases, they become harder.

Over the last few years, I've worked on Rails SaaS platforms serving thousands of users.

The biggest challenges were rarely feature development.

They were:
* Scalability
* Reliability
* Observability
* Operational complexity

So my view is slightly different.

I don't think SaaS is dead. I think SaaS is evolving.

The ability to build software is becoming cheaper. The ability to operate software reliably at scale is becoming more valuable.

And that's where architecture, DevOps, and engineering discipline still matter.

Curious how others see it.
Is SaaS dying, or simply changing shape?