I just did something that most founders would call insane. I moved my entire SaaS business, a platform with thousands of paying customers and millions in revenue, to a completely new cloud AI provider.
It was the single most terrifying decision of my career. A single misstep could have brought the whole thing crashing down. We’re talking about a multi-million dollar business, my reputation, and the jobs of my entire team on the line. So why on earth would I do it?
It started with a bill. A huge, eye-watering bill from our cloud provider. We were scaling fast, which is a good problem to have, but our infrastructure costs were scaling even faster. It felt like we were pouring money into a black hole. I’d lie awake at night wondering if our business model was fundamentally broken. Was I building a company or just a very expensive hobby for my cloud provider?
The Ticking Time Bomb of AI Infrastructure Costs
Let’s be real for a second. The dirty secret of the AI boom is that for many companies, the economics are just brutal. You’re paying an arm and a leg for GPU instances, and the pricing models are often opaque and unpredictable. We were locked into a long-term contract with one of the big three cloud providers, and it felt like we were on a treadmill we couldn’t get off.
Our margins were getting squeezed with every new customer. We were building amazing features, our customers loved the product, but the financial model was unsustainable. I knew we had to make a change, but the thought of a full migration was paralyzing. It’s the kind of thing that keeps you up at night. The kind of thing that can kill a company.
I’ve been in this game for a while. I’ve had two successful exits, RemoteTeam and MovieLaLa, and I’ve invested in over 200 companies, including some of the biggest names in AI like Anthropic, OpenAI, and Scale AI. I’ve seen what it takes to win, and I knew that we couldn’t win if we were constantly being bled dry by our infrastructure costs.
The Search for a Better Way
So, I started looking for alternatives. I talked to dozens of founders, engineers, and investors. I explored every option I could find, from on-premise solutions to smaller, more specialized cloud providers. I was looking for a partner, not just a vendor. I wanted a company that was as obsessed with AI as we were, a company that was building the future, not just renting out servers.
That’s when I found a new player in the space. A company that was offering a serverless AI platform with a pricing model that was so good, I thought it had to be a typo. They were promising to cut our costs by 50-70% while improving performance. It sounded too good to be true.
I was skeptical, to say the least. I’ve seen a lot of hype in my career, and I’ve learned to be wary of anyone who promises the world. But I was also desperate. So, I decided to take a closer look. I had my team run a small-scale test, and the results were mind-blowing. The performance was better, the costs were lower, and the developer experience was a dream.
The Point of No Return
Now I had a decision to make. Do I stick with the devil I know, or do I take a leap of faith on this new provider? It was a classic innovator’s dilemma. The safe bet was to stay put, to keep tinkering with our existing setup, to try and squeeze out a few more percentage points of efficiency. But I knew that wasn’t going to be enough. We needed a step-change, not an incremental improvement.
I called a meeting with my leadership team. I laid out the numbers, the risks, and the potential rewards. I told them that I wanted to go all-in. I wanted to migrate the entire platform. There was a long silence. Then, my CTO, a brilliant engineer who I’ve worked with for years, said, “It’s going to be hell. But I think we have to do it.”
That was all I needed to hear. We spent the next three months planning the migration. We’re talking about a massive undertaking. We had to rewrite huge chunks of our codebase, migrate petabytes of data, and test everything to within an inch of its life. We had to do all of this without our customers noticing a thing. It was like changing the engine of a plane while it was in mid-flight.
The Migration from Hell (and Back)
The migration itself was a 48-hour marathon of caffeine, code, and cold sweats. We had a war room set up in the office, with dashboards tracking every metric imaginable. We had a team of engineers on standby, ready to jump on any issue that came up. I’ve never been so stressed in my life.
There were a few hairy moments. A database that refused to sync. A caching layer that decided to go on strike. But the team was incredible. They were like a well-oiled machine, diagnosing and fixing problems in real-time. I’ve never been more proud of a group of people in my life.
And then, after 48 hours of non-stop work, it was done. We had migrated the entire platform. I remember the moment we flipped the switch. The whole team was holding their breath. And then… nothing happened. No alarms. No angry customer emails. Just the quiet hum of our new, more efficient infrastructure.
The Aftermath: Was It Worth It?
So, was it worth it? The short answer is: hell yes. Our infrastructure costs are down by over 60%. Our performance is up by 30%. And our developers are happier than they’ve ever been. We’re now able to ship features faster, experiment more, and focus on what we do best: building a great product.
But the biggest benefit of the migration wasn’t the cost savings or the performance improvements. It was the change in mindset. We’re no longer afraid to make big, bold bets. We’re no longer held hostage by our infrastructure. We’re in control of our own destiny.
I know that a full migration isn’t for everyone. It’s a high-risk, high-reward move. But if you’re a founder who’s feeling the pain of rising AI infrastructure costs, I would urge you to at least explore the alternatives. Don’t be afraid to challenge the status quo. Don’t be afraid to take a leap of faith. It might just be the best decision you ever make.
I’m not going to tell you who we migrated to. That’s not the point of this post. The point is that you have options. You don’t have to be a slave to the big three cloud providers. There’s a whole new generation of AI infrastructure companies that are building amazing things. Go find them. And if you’re lucky, you might just find a partner who can help you build the future.
What’s the craziest thing you’ve ever done to save your company? I want to hear about it. Leave a comment below. Let’s start a conversation. Because in this crazy, ever-changing world of AI, the only thing we can count on is each other.
Frequently Asked Questions
How can I apply this thinking to my own situation?
Start by identifying the core principle behind the opinion, not the specific example. Then ask yourself: does this principle apply to my context? If yes, test it in a small, low-risk way before going all in.
Do all experts agree with this view?
No, and that's fine. The best ideas in business are often contrarian. I share my perspective based on my experience and data, but I encourage you to seek out opposing viewpoints and form your own conclusions.
What's the most common pushback you get on this?
People often push back by citing exceptions or edge cases. And they're usually right that exceptions exist. But building a strategy around exceptions rather than patterns is a losing game for most founders.
What experience informs this perspective?
This perspective comes from over a decade of building companies in Silicon Valley, two successful exits (RemoteTeam to Gusto, MovieLaLa to Gfycat), and investing in 200+ startups including Anthropic, OpenAI, and Scale AI. I write about what I've lived.